Reno v. ACLU Findings of Fact: Difference between revisions

From Bibliotheca Anonoma
(add shea)
Line 7: Line 7:
These comprehensive findings of fact were made by this court of 3 judges. They are here republished because they provide a detailed overview of what the Internet was like in 1996. As a work of the US government, it is not copyrighted. Page numbers and unrelated footnotes have been removed, but no other changes have been made.
These comprehensive findings of fact were made by this court of 3 judges. They are here republished because they provide a detailed overview of what the Internet was like in 1996. As a work of the US government, it is not copyrighted. Page numbers and unrelated footnotes have been removed, but no other changes have been made.


=FINDINGS OF FACT=
=FINDINGS OF FACT (ACLU)=
All parties agree that in order to apprehend the legal questions at issue in these cases, it is necessary to have a clear understanding of the exponentially growing, world-wide medium that is the Internet, which presents unique issues relating to the application of First Amendment jurisprudence and due process requirements to this new and evolving method of communication. For this reason all parties insisted on having extensive evidentiary hearings before the three-judge court. The court's Findings of fact are made pursuant to Fed.R.Civ.P. 52(a). The history and basic technology of this medium are not in dispute, and the first forty-eight paragraphs of the following Findings of fact are derived from the like-numbered paragraphs of a stipulation the parties filed with the court.
All parties agree that in order to apprehend the legal questions at issue in these cases, it is necessary to have a clear understanding of the exponentially growing, world-wide medium that is the Internet, which presents unique issues relating to the application of First Amendment jurisprudence and due process requirements to this new and evolving method of communication. For this reason all parties insisted on having extensive evidentiary hearings before the three-judge court. The court's Findings of fact are made pursuant to Fed.R.Civ.P. 52(a). The history and basic technology of this medium are not in dispute, and the first forty-eight paragraphs of the following Findings of fact are derived from the like-numbered paragraphs of a stipulation the parties filed with the court.
==The Nature of Cyberspace==
==The Nature of Cyberspace==
Line 495: Line 495:


123. The plaintiffs in this action are businesses, libraries, non-commercial and not-for-profit organizations, and educational societies and consortia. Although some of the material that plaintiffs post online such as information regarding protection from AIDS, birth control or prison rape is sexually explicit and may be considered "indecent" or "patently offensive" in some communities, none of the plaintiffs is a commercial purveyor of what is commonly termed "pornography."
123. The plaintiffs in this action are businesses, libraries, non-commercial and not-for-profit organizations, and educational societies and consortia. Although some of the material that plaintiffs post online such as information regarding protection from AIDS, birth control or prison rape is sexually explicit and may be considered "indecent" or "patently offensive" in some communities, none of the plaintiffs is a commercial purveyor of what is commonly termed "pornography."
=FINDINGS OF FACT (Shea)=
We enter the following findings of fact, many of which are undisputed, the subject of stipulations by the parties, or submitted by the defendant and adopted by us, pursuant to Rule 52(a) of the Federal Rules of Civil Procedure. Although we here consider a so-called facial challenge to a statute, we deemed it appropriate and necessary in the unusual circumstances presented here, and a reasonable exercise of our discretion, to establish a basic record of the facts regarding the new and evolving communications media that is the subject of this legislation.
Section 223(d) targets the use of an “interactive computer service” to send or display patently offensive materials. Although § 223 itself contains no definition of that term, the definition applicable to the new 47 U.S.C. § 230—also added by the CDA—makes clear that the term encompasses means of making “content”<ref>We use the term “content” to refer to any text, data, sound, program, or visual image transmitted over or made available for retrieval on an interactive computer service.</ref> available to multiple users both on the vast web of linked networks popularly known as “the Internet” and on other information systems (such as electronic bulletin boards maintained by educational institutions or nonprofit organizations) not physically linked to the Internet. See Pub.L. No. 104–104, § 509(e)(2), 110 Stat. at 139 (to be codified at 47 U.S.C. § 230(e)(2)). We draw upon the stipulations of the parties and the testimony adduced at the three-day evidentiary hearing to describe: (1) the nature of the medium targeted by § 223(d), focusing in part on the degree of control that those who transmit content have over who will receive it; (2) the availability of certain categories of potentially objectionable material on line; (3) the development of software and labeling standards enabling parents to limit their children's exposure to objectionable on-line content; and (4) the potential for tagging and verification procedures that content providers can use in an effort to shield minors from sexually explicit content that they provide.<ref>While § 223(d) regulates more than the content of Internet communications, we focus mainly on the range of tools and services available to individuals with Internet access, recognizing that the vast majority of content available through the use of an interactive computer service is in fact available on the Internet.</ref> As we do so, we unavoidably—and with apologies to all others with a similar aversion to “cyberspeak”—adopt some of the terminology that has developed in conjunction with this technology. We endeavor, to the extent possible, to avoid the jargon of this field, and to define our terms wherever possible, for the sake of the clarity of this record and this opinion, as well as for the benefit of any reader required to review our work.
==A. The Development of the Internet==
Although “the Internet” now formally describes a collection of more than 50,000 networks linking some nine million host computers in ninety countries, it has existed for nearly three decades on a much smaller scale. What we now refer to as the Internet grew out of an experimental project of the Department of Defense's Advanced Research Projects Administration (“ARPA”) designed to provide researchers with direct access to supercomputers at a few key laboratories and to facilitate the reliable transmission of vital communications. (Declaration of William J. Hoffman (“Hoffman Decl.”), Ex. 4, at 11–12) ARPA supplied funds to link computers operated by the military, defense contractors, and universities conducting defense-related research through dedicated phone lines, creating a “network” known as ARPANet. (Parties' Stipulations in Preparation for Preliminary Injunction Hearing (“Joint Stip.”) ¶¶ 6–7; Hoffman Decl., Ex. 3, at 3; id. Ex. 4, at 11) Programs on the linked computers implemented a technical scheme known as “packet-switching,” through which a message from one computer to another would be subdivided into smaller, separately addressed pieces of data, known as “packets,” sent independently to the message's destination and reassembled upon arrival. (Joint Stip. ¶ 9) Each computer on the network was in turn linked to several other computers, creating any number of routes that a communication from one computer could follow to reach its destination. If part of the network were damaged, a portion of the message could be re-routed automatically over any other path to its ultimate destination, a characteristic of the network intended initially to preserve its operability in the event of enemy attack. (Id. ¶¶ 7–8; Hoffman Decl., Ex. 3, at 3; id. Ex. 4, at 12)
Having successfully implemented a system for the reliable transfer of information over a computer network, ARPA began to support the development of communications protocols for transferring data between different types of computer networks. Universities, research facilities, and commercial entities began to develop and link together their own networks implementing these protocols; these networks included a high-speed “backbone” network known as NSFNet, sponsored by the National Science Foundation, smaller regional networks, and, eventually, large commercial networks run by organizations such as Sprint, IBM, and Performance Systems International (commonly known as “PSI”). (Hoffman Decl., Ex. 3, at 3; id. Ex. 4, at 13–14) As faster networks developed, most network traffic shifted away from ARPANet, which formally ceased operations in 1990. (Id. Ex. 3, at 3) What we know as “the Internet” today is the series of linked, overlapping networks that gradually supplanted ARPANet. Because the Internet links together independent networks that merely use the same data transfer protocols, it cannot be said that any single entity or group of entities controls, or can control, the content made publicly available on the Internet or limits, or can limit, the ability of others to access public content. Rather, the resources available to one with Internet access are located on individual computers around the world. (Joint Stip. ¶ 11)
It is estimated that as many as forty million individuals have access to the information and tools of the Internet, and that figure is expected to grow to 200 million by the year 1999. (Id. ¶ 3) Access to the Internet can take any one of several forms. First, many educational institutions, businesses, libraries, and individual communities maintain a computer network linked directly to the Internet and issue account numbers and passwords enabling users to gain access to the network directly or by modem.<ref>A “modem” (a contraction of “modulator” and “demodulator”) is a device that translates digital information into a signal for transmission over a telephone line (“modulation”) and translates a signal received over a telephone line into digital information (“demodulation”).</ref> (Id. ¶¶ 12–14) Second, “Internet service providers,” generally commercial entities charging a monthly fee, offer modem access to computers or networks linked directly to the Internet. (Id. ¶ 16) Third, national commercial “on-line services”—such as America Online, CompuServe, Prodigy, and Microsoft Network—allow subscribers to gain access to the Internet while providing extensive content within their own proprietary networks. (Id. ¶ 17) Finally, organizations and businesses can offer access to electronic bulletin-board systems—which, like national on-line services, provide certain proprietary content; some bulletin-board systems in turn offer users links to the Internet. (Id. ¶ 18)
==B. Categories of Internet Use==
For our purposes, there are two loose and overlapping categories of Internet use. First, an individual who has secured access to the Internet can correspond or exchange views with one or many other Internet users. Second, a user can locate and retrieve information available on other computers. We explore these categories in greater detail below. As will become clear, distinctions in how Internet content is transmitted affect the degree of control that providers of content have over who will be able to gain access to their communications;<ref>We use the term “content provider” to refer to any Internet “speaker”—that is, a user who transmits or makes available any content over the Internet. Although the term is not used in the statutory provision at issue, “information content provider” is elsewhere defined in the CDA as “any person or entity that is responsible, in whole or in part, for the creation or development of information provided through the Internet or any other interactive computer service.” Pub.L. No. 104–104, § 509(e)(3), 110 Stat. at 139 (to be codified at 47 U.S.C. § 230(e)(3)). The phrase serves as a reasonable shorthand for the category of individuals targeted by the CDA—persons who send or display Internet content. See S.Conf.Rep. No. 230, 104th Cong., 2d Sess. 188 (1996) (stating that § 223(d)(1) applies to “content providers”).</ref> we will return to the legal significance of these distinctions at a later juncture. For any communication to take place over the Internet, two pieces of software,<ref>We use the term “software” to refer to the combination of programs and procedures that serve as instructions to the computer. The term is often used in contrast with “hardware,” which refers to a computer system's physical elements.</ref> adhering to the same communications protocol, are required. A user must have access to certain kinds of “client” software, which enables his computer to communicate with and make requests of remote computers where information is stored; these remote computers must be running “server” software, which provides information in response to requests by client software. (Declaration of Dr. Dan R. Olsen, Jr. (“Olsen Decl.”), ¶¶ 13–14)
===1. Communicating with Other Internet Users===
Perhaps the most widely used Internet service is electronic mail, or “e-mail.” Using any one of dozens of available “mailers”—client software capable of reading and writing e-mail—a user is able to address and transmit a message to one or more specific individuals. (Joint Stip. ¶ 21) A user can also “subscribe” to an electronic mailing list on a topic of interest; the user receives a copy of messages posted by other subscribers and, in turn, can post messages for forwarding to the full mailing list. Once a mailing list is established, it is typically maintained using a “mail exploder”—a program such as “listserv” running on the server on which the list resides—that automatically (i.e., without human intervention) responds to a user's request to be added to or removed from the list of subscribers and retransmits messages posted by a subscriber to others on the mailing list. (Id. ¶ 22) Some mailing lists are “closed”: a user's request to join the list requires the approval of an individual who maintains the list. (Id.) Mailing lists (both open and closed) may also be “moderated”: all messages posted to the list are forwarded to a moderator, who approves certain messages and retransmits them to subscribers. (Id.) An individual sending a message that will be retransmitted by a mail exploder program has no way of knowing the e-mail addresses of other subscribers. (Olsen Decl. ¶ 19; Testimony of Gordon C. Galligher, Jr., Tr. at 181) Even if the user could obtain an e-mail address for each subscriber to a particular list, those addresses alone would provide no authoritative information about subscribers. There is no directory that identifies persons using a certain e-mail address. In addition, a user can avoid disclosing his true e-mail address by developing an e-mail “alias” or by using an “anonymous remailer”—a server that purges identifying information from a communication before forwarding it to its destination. (Defendant's Response to Plaintiff's Request for Admissions (“Defendant's Adm.”) No. 22; Galligher Test., Tr. at 173)
Internet users may also transmit or receive “articles” posted daily to thousands of discussion groups, arranged by subject matter and known as “newsgroups,” available through an electronic bulletin-board system known as “Usenet.” When a user with access to a Usenet server—that is, a computer participating in the Usenet system—posts an article to a particular newsgroup, the server automatically forwards the article to adjacent Usenet servers, which in turn forward it to other servers, until the article is available on all Usenet sites that furnish access to the newsgroup in question. (Joint Stip. ¶ 23) Once a message reaches a particular Usenet site, it is temporarily stored there so that individual users—running client software, known as a “newsreader,” capable of sorting articles according to header information identifying the newsgroup to which the article was posted—can review and respond to the message. (Id.; Hoffman Decl., Ex. 4, at 129) Some Usenet newsgroups are moderated; messages to the newsgroup are forwarded to an individual who selects those appropriate for distribution. (Joint Stip. ¶ 23) Because Usenet articles are distributed to (and made available on) multiple servers, one who posts an article to a newsgroup has no way of knowing who will choose to retrieve it, whether or not the newsgroup is moderated. (Galligher Test., Tr. at 170, 174–75) There is no newsgroup equivalent of a “closed” mailing list: access to a particular newsgroup can only be limited by restricting the number of servers participating in the newsgroup. (Testimony of Clay Shirky, Tr. at 251)
The Internet also offers opportunities for multiple users to interact in real time. Using a program called “Talk,” two users can exchange messages while they are both on line; a message typed on one user's computer will appear almost immediately on the other's screen. (Joint Stip. ¶ 25) Servers running so-called “chat” software, such as Internet Relay Chat (“IRC”), permit multiple users to converse by selecting one of many discussion “channels” active at any time. Commercial on-line services such as America Online, CompuServe, Prodigy, and the Microsoft Network offer their own chat systems for their members. (Id. ¶ 26) Having joined a channel, the user can see and read messages transmitted by other users, each identified by a name the user selects upon joining the channel. (Id. ¶ 25) Individual participants in IRC discussions know other participants only by the names they choose upon entering the discussion; users can participate anonymously by using a pseudonym.
===2. Locating and Retrieving Information on the Internet===
Individuals with Internet access can take advantage of a number of tools for locating and retrieving information and resources stored on remote computers. One who wishes to make certain articles, files, or software available to other users will set up a server, adhering to certain communications protocols, capable of retrieving and presenting stored information in response to a request from client software using the same communications protocol. (Olsen Decl. ¶¶ 13, 16; Galligher Test., Tr. at 131)
====a. File–Transfer Protocol (“FTP”)====
One type of software implements a set of conventions for copying files from a host computer known as “file-transfer protocol” (“FTP”). With appropriate client software, a user with an account on the host computer can contact the server, view a directory of available files, and copy one or more of those files to his own computer. In addition to making files available to users with accounts, thousands of content providers also make files available for “anonymous” retrieval by users who do not possess an account on the host computer.<ref>To locate files available for copying, a user can contact an “Archie” server—a remote computer capable of searching directories for file names containing a particular string of characters on FTP servers permitting anonymous retrieval. (Hoffman Decl., Ex. 4, at 180–90) (Hoffman Decl., Ex. 3, at 1–2, 5; id. Ex. 4, at 187; Joint Stip. ¶ 29)</ref> A content provider who makes files available for retrieval by anonymous FTP has no way of discerning who gains access to the files.
====b. “Gopher ” Servers====
A second type of server software capable of making available the resources of a host computer is known as a “gopher” program. (Joint Stip. ¶ 30, Hoffman Decl., Ex. 3, at 5) A gopher server presents information in a set of menus, enabling a user who gains access to the server to select a series of increasingly narrow menu items before locating a desired file that can be displayed on or copied to the user's computer.<ref>As with FTP servers, there are tools available for locating menus or items containing a certain string of characters: a “Veronica” server is capable of searching menus on all gopher servers, while “Jughead” is an aptly named tool for searching menus on only a single server. (Galligher Test., Tr. at 124; Hoffman Decl., Ex. 3, at 5; id. Ex. 4, at 191–92)</ref> (Galligher Test., Tr. at 122; Hoffman Decl., Ex. 3, at 5) A content provider who maintains a gopher server ordinarily has no way of knowing who will gain access to the information made available.
====c. The World Wide Web====
The third and perhaps best known method of locating and accessing information on the Internet is by exploring the World Wide Web. Documents available on the Web are not collected in any central location; rather, they are stored on servers around the world running Web server software. (Joint Stip. ¶¶ 31, 38, 40) To gain access to the content available on the Web, a user must have a Web “browser”—client software, such as Netscape Navigator, Mosaic, or Internet Explorer, capable of displaying documents formatted in “hypertext markup language” (“HTML”), the standard Web formatting language. (Galligher Test., Tr. at 125; Joint Stip. ¶¶ 31, 43) Each document has an address, known as a Uniform Resource Locator (“URL”), identifying, among other things, the server on which it resides; most documents also contain “links”—highlighted text or images that, when selected by the user, permit him to view another, related Web document. (Joint Stip. ¶ 34) Because Web servers are linked to the Internet through a common communications protocol, known as hypertext transfer protocol (“HTTP”), a user can move seamlessly between documents, regardless of their location; when a user viewing a document located on one server selects a link to a document located elsewhere, the browser will automatically contact the second server and display the document. (Joint Stip. ¶¶ 34, 37) Some types of Web client software also permit users to gain access to resources available on FTP and gopher sites.
A number of “search engines”—such as Yahoo, Magellan, Alta Vista, WebCrawler, and Lycos—are available to help users navigate the World Wide Web.<ref>Most of these services do not charge users for search requests and are sustained primarily by advertising revenues. (Galligher Test., Tr. at 136–37)</ref> For example, the service Yahoo maintains a directory of documents available on various Web servers. A user can gain access to Yahoo's server and type a string of characters as a search request. Yahoo returns a list of documents whose entries in the Yahoo directory match the search string and organizes the list of documents by category. (Galligher Test., Tr. at 134; Plaintiff's Ex. 3) Search engines make use of software capable of automatically contacting various Web sites and extracting relevant information. Some search engines, such as Alta Vista, store the information in a database and return it in response to a user request. Others, such as Yahoo, employ a group of individuals to determine whether and how a site should be categorized in the Yahoo directory. (Galligher Test., Tr. at 137; Supplemental Declaration of William J. Hoffman (“Hoffman Supp.Decl.”) Ex. A, at 39–42 (Testimony of Donna L. Hoffman in ACLU/ALA ))
As the growth in Internet use and the wide availability of tools and resources to those with access to the Internet suggest, the Internet presents extremely low entry barriers to those who wish to convey Internet content or gain access to it. In particular, a user wishing to communicate through e-mail, newsgroups, or Internet Relay Chat need only have access to a computer with appropriate software and a connection to the Internet, usually available for a low monthly fee. The user then in a sense becomes a public “speaker,” able to convey content, at relatively low cost, to users around the world to whom it may be of interest. Those who possess more sophisticated equipment and greater technical expertise can make content available on the Internet for retrieval by others (known or unknown) by running a server supporting anonymous FTP, a gopher server, or a Web server. Yet content providers need not necessarily run their own servers or have the programming expertise to construct their own sites; they can lease space on a Web server from another or create a “home page” through an on-line commercial service.
The ease of entry of many speakers sets interactive computer systems apart from any other more traditional communications medium that Congress has attempted to regulate in the past. With one-way media such as radio and television broadcasting or cable programming, a user is merely a listener or viewer; in the CDA, Congress sought to target “interactive” computer systems through which a listener or viewer, by definition, has the power to become a speaker. The relative ease of speaker entry and the relative parity among speakers accounts for the unprecedented and virtually unlimited opportunities for political discourse, cultural development, and intellectual activity that Congress found to characterize emerging communication technologies.
In seeking to describe the range of tools and opportunities for Internet users to “speak,” we recognize that the categories we delineate are far from clean and the technology is far from static. Indeed, by all indications, the way that we conceptualize various media that we have traditionally viewed as distinct—such as cable television, telephones, and computer networks—will change dramatically as these media “converge” into common forms of communication. See Denver Area Educ. Telecommunications Consortium v. FCC (“Denver Area Consortium”), 518 U.S. 727, –––– & n. 4, 116 S.Ct. 2374, 2402 & n. 4, 135 L.Ed.2d 888 (1996) (Souter, J., concurring); see also Jerry Berman & Daniel J. Weitzner, Abundance and User Control: Renewing the Democratic Heart of the First Amendment in the Age of Interactive Media, 104 Yale L.J. 1619, 1619 n. 1 (1995); Art Kramer, Netwatch: The AJC's Daily Online Guide, Atl.J. & Const., May 29, 1996, at B04 (describing cable modem technology designed to offer Internet access through existing cable television connections) (Hoffman Supp.Decl., Ex. C, at 3–4). Of course, our findings of fact are necessarily time-bound. We can only determine whether the statutory provision at issue here, in light of the technology available during the pendency of this case, comports with the First Amendment.
==C. Sexually Explicit Content on the Internet==
It is undisputed that there exists some content on the Internet that is—to use the Government's phrase—“sexually explicit.” (Defendant's Memorandum of Law, filed March 19, 1996, at 11) The term “sexually explicit” is descriptive rather than legal and does not appear in the statutory provision at issue, but the Government employs it as a shorthand to describe Internet content depicting “sexual or excretory activities or organs”—possibly though not necessarily in a patently offensive way. (Defendant's Supplemental Memorandum of Law (“Defendant's Supp. Memo.”), filed June 7, 1996, at 9) That is, the Government does not contend that all sexually explicit material is “patently offensive” and therefore within the scope of the CDA, but claims that there is certainly content available on the Internet that is both sexually explicit and patently offensive.
The testimony and demonstration of one of the Government's expert witnesses, Howard Schmidt, Director of the Air Force Office of Special Investigations, amply confirmed the availability of sexually explicit material on line. Nevertheless, there is no persuasive evidence in the record to suggest, much less prove, that sexually explicit material easily “assaults” an unknowing user—as in other media, most notably television and radio—or that any substantial proportion of Internet content is sexually explicit.
===1. Ease of Access to Sexually Explicit Content===
The Government urges us to conclude that an Internet user can easily stumble upon sexually explicit material. (Defendant's Post–Hearing Memorandum of Law (“Defendant's Post–Hearing Memo.”), filed May 28, 1996, at 31–32) It is important to begin with the general observation that, with the exception of e-mail, no content appears on a user's screen without the user having first taken some affirmative step. One wishing to read articles posted to a newsgroup must connect to a Usenet server and select the relevant group. To retrieve a file through anonymous FTP or access a gopher server, the user must search for or know the address of a particular server. To gain access to content on the World Wide Web, a user must know the URL of a relevant site or type a keyword into one of several available search engines.
Schmidt's demonstration focused mainly on the availability of sexually explicit content on the World Wide Web. In the absence of any screening software or filter, a user determined to view a site containing sexually explicit material can certainly do so, either by typing a known URL or by searching for key words. One sexually explicit site may, in turn, contain “links” to other such sites. (Defendant's Exs. 13, 16, 17, 26, 29, 32; Schmidt Test., Tr. at 401–02) While ordinarily a user must affirmatively seek sexually explicit material to view it, on occasion a search not intended to retrieve sexually explicit material may retrieve a link to a sexually explicit site. For example, Schmidt's searches of “Sleeping Beauty,” “Babe,” and “Little Women” produced a handful of links to sexually explicit sites. (Defendant's Exs. 15, 18, 27, 31, 38) This demonstration revealed the inevitable imprecision of search engines—a broad search will almost always return some irrelevant results. In the vast majority of cases, the character of a sexually explicit site will be clear from the entry or link that a search engine returns. Nevertheless, there is potential for occasional accidental viewing of sexually explicit material. For example, if a user were to view entries in a WebCrawler search using that program's standard format as preset by the manufacturer, he would see no summary of the sites' contents. (Defendant's Ex. 18; Shirky Test., Tr. at 237–38) One of Schmidt's searches of “Sleeping Beauty” returned an entry offering a link to a site containing sexually explicit material; the entry (when viewed apart from other entries on the same page with similar addresses) gave little indication of the site's contents. (Defendant's Ex. 15; Shirky Test., Tr. at 238) It is difficult to know how often accidental viewing can occur, but there is no basis in the record for concluding that a user not seeking out sexually explicit material on the Internet will encounter it with any particular frequency.
===2. The Availability of Sexually Explicit Content===
Although Schmidt's demonstration focused on the World Wide Web, sexually explicit content is available on the Internet through almost any form of Internet communication. Yet there is no evidence that sexually explicit content constitutes a substantial—or even significant—portion of available Internet content. While it is difficult to ascertain with any certainty how many sexually explicit sites are accessible through the Internet, the president of a manufacturer of software designed to block access to sites containing sexually explicit material testified in the Philadelphia litigation that there are approximately 5,000 to 8,000 such sites, with the higher estimate reflecting the inclusion of multiple pages (each with a unique URL) attached to a single site. (Stipulated Portions of Record in ACLU/ALA (“Stipulated Record”), Ex. M, at 139–40 (Testimony of Ann W. Duvall in ACLU/ALA )) The record also suggests that there are at least thirty-seven million unique URLs. (Galligher Test., Tr. at 144) Accordingly, even if there were twice as many unique pages on the Internet containing sexually explicit materials as this undisputed testimony suggests, the percentage of Internet addresses providing sexually explicit content would be well less than one tenth of one percent of such addresses.
It is not disputed that some of the sexually explicit materials that the CDA attempts to keep away from minors originates abroad. This is not surprising inasmuch as forty percent of all host computers are located outside the United States. (Joint Stip. ¶ 3) Although only a tentative approximation is possible, the record suggests that as much as thirty percent of the sexually explicit material currently available on the Internet originates in foreign countries. (Stipulated Record, Ex. L, ¶ 41; id. Ex. M, at 161–62 (Duvall Test.))
==D. The Development of Blocking Tools and Labeling Schemes==
As the Internet has become accessible to more households, several commercial on-line services and software companies have developed features and packages designed to enable parents to limit children's exposure to potentially inappropriate Internet material. *932 For example, America Online, Prodigy, and Microsoft Network, which permit their subscribers to obtain access to Internet material, offer parental control options free of charge to their members. (Joint Stip. ¶ 67) America Online, for example, allows parents to establish a separate account for their children limited to the service's own proprietary content. (Id.) In addition, at least one type of screening software, SurfWatch, has a feature allowing parents to block access to all Internet sites except for those that parents choose to make available to their children. (Stipulated Record, Ex. M, at 131 (Duvall Test.))
The Government offered testimony and a demonstration regarding SurfWatch (configured to act as a screening tool, rather than to block all Internet access) and a second type of screening software, Cyber Patrol. SurfWatch and Cyber Patrol maintain lists of sites known to contain sexually explicit material; when operating while a user attempts to retrieve Internet material, access to sites identified on their programs will be blocked. In addition, the programs block access to sites whose URLs contain particular character patterns or words, such as “xxx” or “sex,” and block any searches including those character patterns or words.
Because of the constant change in the number and location of Internet sites, both SurfWatch and Cyber Patrol offer regular subscription or update services. But even where a parent has properly installed screening software and the software is operational (and configured to block access to certain sites rather than to the entire Internet), it is possible to retrieve some sexually explicit material. The Government's witness was able to run searches using “Babe” and “Little Women” as key words with screening software running in the background. As with searches performed in the absence of screening software, the searches returned links to sexually explicit materials. Some of the links were not blocked by the screening tool. In addition, the Government's witness obtained access to sexually explicit material by directly entering URLs obtained from earlier searches conducted without blocking software in the background. The record also shows that blocking software is not widely owned by or used in households with access to the Internet: nearly seventy percent of SurfWatch's 1,500 subscribers are schools rather than individual households. (Id. at 163–65)
Other efforts to assist parents in filtering and screening material that their children can view on the Internet are under way. The World Wide Web Consortium (“W3C”) has launched the Platform for Internet Content Selection (“PICS”) to develop technical standards for attaching electronic ratings to Internet addresses. (Joint Stip. ¶¶ 47–49; Stipulated Record, Ex. J., at 1; id. Ex. G, at 2–3 (Declaration of Albert Vezza in ACLU/ALA )) When the system is fully implemented, PICS-compatible client software (including browsers, newsgroup readers, and mail readers); Internet service providers; and commercial on-line services will be able to detect PICS tags and block content based on how a parent has configured the software. (Joint Stip. ¶ 48; Stipulated Record, Ex. G., at 3 (Vezza Decl.)) PICS will thus enable parents to design from an array of categories blocking criteria that suit the parents' values or needs. The PICS program envisages both rating by content providers and rating by third parties. (Joint Stip. ¶ 48) The vast majority of Internet sites currently remain unrated. Nevertheless, Microsystems Software, Inc. (which manufactures Cyber Patrol) introduced a PICS ratings server in February 1996. (Id. ¶ 54) Cyber Patrol is itself now PICS-compatible; it can screen out material based on its PICS tag. (Id.) In addition, Microsoft released the first PICS-compatible Web browser, Internet Explorer 3.0, on May 28, 1996. The browser allows parents to block children's access to all unrated Internet sites and to specify appropriate levels of violence or nudity at rated sites. (Hoffman Supp.Decl., Ex. C, at 1–3)
In addition to PICS tags, the Government's expert witness, Dr. Dan Olsen, testified that content providers wishing to transmit or make available material potentially falling within the scope of the CDA could develop a general practice of inserting a “tag” or “label”—a string of characters, such as “–L18” (for “not less than 18 years”)—into the address or name of a particular site so as to clearly identify the site as unsuitable for minors. To transmit or gain access to Internet content, a user must specify a textual name: one cannot send e-mail without an e-mail address or the name of a mailing list; post an article to a newsgroup without specifying the name of the group; participate in the Internet Relay Chat without specifying a “channel”; or access a file without its address. (Olsen Decl. ¶¶ 22–26) Accordingly, content providers using all significant modes of Internet communication could use a tag to identify their content as “covered” content. For example, when a sender transmits an e-mail message, the message is accompanied by the sender's address, which contains a “user name” identifying a particular user and a “domain name” assigned to a computer or set of computers.<ref>In the example [email protected], “smith.com” would constitute a domain name.</ref> (Olsen Decl. ¶¶ 25, 60) If the string –L18 were added to the domain name, all e-mail originating from that site—regardless of the particular user who transmitted it—would be identified as containing material falling within the scope of the CDA.<ref>Following the example above, all e-mail would originate from the domain smith-L18.com.</ref> In the alternative, a particular user name—rather than a domain name—could contain the “–L18” tag; only e-mail originating under that user name would be tagged.<ref>In the example above, material would originate from the address [email protected].</ref> Finally, a tag could be placed in a textual subject line, so as to identify only particular messages (rather than all e-mail sent under a certain user name or from a certain computer) as containing content potentially within the scope of the CDA. (Id. ¶¶ 60–62)
Similarly, a tag such as “–L18” could be added to the name of a newsgroup; an individual user wishing to post an article potentially falling within the scope of the CDA to a newsgroup that does not as a general matter contain such material could insert a tag in the subject line accompanying the article. (Id. ¶¶ 64–65) A tag could also be placed in the name of an IRC channel.
Turning to means of making files available for retrieval or viewing by remote users—using an FTP, gopher, or Web server—content providers could insert a specific tag such as “–L18” in a domain name or site name. Thus, as the Government's expert witness testified, an owner of a Web site named “www.cyberporn.com” could rename the site “www–L18.cyberporn.com”. (Id. ¶ 51) If a site only contained specific files falling within the scope of the CDA, a content provider could identify those files by adding a tag to the name of the directory in which the file resides or to the file name itself. That is, a file identified with the address “http://www.adult.com/picturel.html/” could be renamed “http://www.adult.com/picture1–L18.html/”; in the alternative, a content provider could place all covered files within a specific directory, such as “http://www.adult.com/pictures–L18/”. (Id. ¶¶ 51–54) A content provider who did not wish to tag an entire file available on a Web server as unsuitable for minors could place a tag within the HTML source code of the file, thus identifying a particular section as subject to the CDA. (Id. ¶ 58) In any of these approaches, tagging content is, in a technical sense, a trivial act. (Id. ¶¶ 59, 62; Stipulated Record, Ex. B, at 56 (Testimony of Scott O. Bradner in ACLU/ALA ))
There is an alternative means to shield minors from sexually explicit content available uniquely to content providers on the World Wide Web: verification of a user's “adulthood” before allowing him access to a site. A content provider operating a Web server can create and display an electronic form to retrieve information from a user visiting the Web site; after processing the information by using a program such as a Common Gateway Interface (“cgi”) script, the server could grant or deny access to the site. (Shirky Decl. ¶ 21) Not all content providers who make material available on the Web, however, can use programs such as cgi scripts; for example, commercial on-line services such as America Online and CompuServe provide subscribers with the opportunity to post content by configuring their own Web pages but do not permit subscribers to use cgi scripts. (Olsen Test., Tr. at 345) For Web content providers who lack access to cgi scripts, there is no means of age verification.
Although some Web providers can query the user of a site for a credit card number, the cost of verification is significant, ranging from sixty cents per transaction to more than a dollar per transaction. (Id. at 341–42) To take advantage of adult access code or adult identification code verification, a content provider would either have to establish and maintain a registration and verification system (or hire someone else to do so) and issue access codes to users—after verifying their ages—or associate with one of several adult verification services, such as Adult Check, Adult Verification System, First Virtual, Validate, or VeriSign. (Olsen Decl. ¶ 86 & Ex. I; Schmidt Test., Tr. at 203–14; Defendant's Exs. 6, 7, 8 & 9) Although neither of the Government's expert witnesses had any first-hand familiarity with adult verification services, advertising materials suggest that an adult can obtain an identification number from a particular service and access any site registered with the service. For example, a user can register with Adult Check for an annual fee of $9.95; when the user attempts to access any site registered with Adult Check, the user is prompted to enter an Adult Check identification number that is checked against the service's database. (Defendant's Ex. 6, at 1) If the number is valid, the user is automatically admitted to the site. (Id.) Although most verification services do not charge content providers to register their sites (Id. Exs. 6–8), at least one service does impose a fee on site owners registered with it. (Id. Ex. 9, at 1)
Having explored various means of Internet communication, the availability and accessibility of sexually explicit content, the development of blocking software and rating schemes designed to enable parents to shield their children from inappropriate material, and the potential for tagging and verification procedures that content providers can themselves employ in an effort to shield minors from sexually explicit content that they provide, we turn to the governmental regulation in question.
==Footnotes==
==Footnotes==
<references />
<references />
[[Category:Law]]
[[Category:Law]]

Revision as of 23:48, 16 October 2021

Introductory Note

Reno v. ACLU

Reno v. ACLU (1997) was the Supreme Court case that assured the survival of the infant Internet by unanimously striking down the Communications Decency Act, which made it illegal to post any "indecent" content anywhere minors could see.

The authors of the CDA, perhaps anticipating its controversy, had provided for an expedited process to review its constitutionality. First, it would be heard by a court of 3 judges. Then, it would go straight up to the Supreme Court which would have the last word.

These comprehensive findings of fact were made by this court of 3 judges. They are here republished because they provide a detailed overview of what the Internet was like in 1996. As a work of the US government, it is not copyrighted. Page numbers and unrelated footnotes have been removed, but no other changes have been made.

FINDINGS OF FACT (ACLU)

All parties agree that in order to apprehend the legal questions at issue in these cases, it is necessary to have a clear understanding of the exponentially growing, world-wide medium that is the Internet, which presents unique issues relating to the application of First Amendment jurisprudence and due process requirements to this new and evolving method of communication. For this reason all parties insisted on having extensive evidentiary hearings before the three-judge court. The court's Findings of fact are made pursuant to Fed.R.Civ.P. 52(a). The history and basic technology of this medium are not in dispute, and the first forty-eight paragraphs of the following Findings of fact are derived from the like-numbered paragraphs of a stipulation the parties filed with the court.

The Nature of Cyberspace

The Creation of the Internet and the Development of Cyberspace

1. The Internet is not a physical or tangible entity, but rather a giant network which interconnects innumerable smaller groups of linked computer networks. It is thus a network of networks. This is best understood if one considers what a linked group of computers referred to here as a "network" is, and what it does. Small networks are now ubiquitous (and are often called "local area networks"). For example, in many United States Courthouses, computers are linked to each other for the purpose of exchanging files and messages (and to share equipment such as printers). These are networks.

2. Some networks are "closed" networks, not linked to other computers or networks. Many networks, however, are connected to other networks, which are in turn connected to other networks in a manner which permits each computer in any network to communicate with computers on any other network in the system. This global Web of linked networks and computers is referred to as the Internet.

3. The nature of the Internet is such that it is very difficult, if not impossible, to determine its size at a given moment. It is indisputable, however, that the Internet has experienced extraordinary growth in recent years. In 1981, fewer than 300 computers were linked to the Internet, and by 1989, the number stood at fewer than 90,000 computers. By 1993, over 1,000,000 computers were linked. Today, over 9,400,000 host computers worldwide, of which approximately 60 percent located within the United States, are estimated to be linked to the Internet. This count does not include the personal computers people use to access the Internet using modems. In all, reasonable estimates are that as many as 40 million people around the world can and do access the enormously flexible communication Internet medium. That figure is expected to grow to 200 million Internet users by the year 1999.

4. Some of the computers and computer networks that make up the Internet are owned by governmental and public institutions, some are owned by non-profit organizations, and some are privately owned. The resulting whole is a decentralized, global medium of communications or "cyberspace" that links people, institutions, corporations, and governments around the world. The Internet is an international system. This communications medium allows any of the literally tens of millions of people with access to the Internet to exchange information. These communications can occur almost instantaneously, and can be directed either to specific individuals, to a broader group of people interested in a particular subject, or to the world as a whole.

5. The Internet had its origins in 1969 as an experimental project of the Advanced Research Project Agency ("ARPA"), and was called ARPANET. This network linked computers and computer networks owned by the military, defense contractors, and university laboratories conducting defense-related research. The network later allowed researchers across the country to access directly and to use extremely powerful supercomputers located at a few key universities and laboratories. As it evolved far beyond its research origins in the United States to encompass universities, corporations, and people around the world, the ARPANET came to be called the "DARPA Internet," and finally just the "Internet."

6. From its inception, the network was designed to be a decentralized, self-maintaining series of redundant links between computers and computer networks, capable of rapidly transmitting communications without direct human involvement or control, and with the automatic ability to re-route communications if one or more individual links were damaged or otherwise unavailable. Among other goals, this redundant system of linked computers was designed to allow vital research and communications to continue even if portions of the network were damaged, say, in a war.

7. To achieve this resilient nationwide (and ultimately global) communications medium, the ARPANET encouraged the creation of multiple links to and from each computer (or computer network) on the network. Thus, a computer located in Washington, D.C., might be linked (usually using dedicated telephone lines) to other computers in neighboring states or on the Eastern seaboard. Each of those computers could in turn be linked to other computers, which themselves would be linked to other computers.

8. A communication sent over this redundant series of linked computers could travel any of a number of routes to its destination. Thus, a message sent from a computer in Washington, D.C., to a computer in Palo Alto, California, might first be sent to a computer in Philadelphia, and then be forwarded to a computer in Pittsburgh, and then to Chicago, Denver, and Salt Lake City, before finally reaching Palo Alto. If the message could not travel along that path (because of military attack, simple technical malfunction, or other reason), the message would automatically (without human intervention or even knowledge) be re-routed, perhaps, from Washington, D.C. to Richmond, and then to Atlanta, New Orleans, Dallas, Albuquerque, Los Angeles, and finally to Palo Alto. This type of transmission, and re-routing, would likely occur in a matter of seconds.

9. Messages between computers on the Internet do not necessarily travel entirely along the same path. The Internet uses "packet switching" communication protocols that allow individual messages to be subdivided into smaller "packets" that are then sent independently to the destination, and are then automatically reassembled by the receiving computer. While all packets of a given message often travel along the same path to the destination, if computers along the route become overloaded, then packets can be re-routed to less loaded computers.

10. At the same time that ARPANET was maturing (it subsequently ceased to exist), similar networks developed to link universities, research facilities, businesses, and individuals around the world. These other formal or loose networks included BITNET, CSNET, FIDONET, and USENET. Eventually, each of these networks (many of which overlapped) were themselves linked together, allowing users of any computers linked to any one of the networks to transmit communications to users of computers on other networks. It is this series of linked networks (themselves linking computers and computer networks) that is today commonly known as the Internet.

11. No single entity academic, corporate, governmental, or non-profit administers the Internet. It exists and functions as a result of the fact that hundreds of thousands of separate operators of computers and computer networks independently decided to use common data transfer protocols to exchange communications and information with other computers (which in turn exchange communications and information with still other computers). There is no centralized storage location, control point, or communications channel for the Internet, and it would not be technically feasible for a single entity to control all of the information conveyed on the Internet.

How Individuals Access the Internet

12. Individuals have a wide variety of avenues to access cyberspace in general, and the Internet in particular. In terms of physical access, there are two common methods to establish an actual link to the Internet. First, one can use a computer or computer terminal that is directly (and usually permanently) connected to a computer network that is itself directly or indirectly connected to the Internet. Second, one can use a "personal computer" with a "modem" to connect over a telephone line to a larger computer or computer network that is itself directly or indirectly connected to the Internet. As detailed below, both direct and modem connections are made available to people by a wide variety of academic, governmental, or commercial entities.

13. Students, faculty, researchers, and others affiliated with the vast majority of colleges and universities in the United States can access the Internet through their educational institutions. Such access is often via direct connection using computers located in campus libraries, offices, or computer centers, or may be through telephone access using a modem from a student's or professor's campus or off-campus location. Some colleges and universities install "ports" or outlets for direct network connections in each dormitory room or provide access via computers located in common areas in dormitories. Such access enables students and professors to use information and content provided by the college or university itself, and to use the vast amount of research resources and other information available on the Internet worldwide.

14. Similarly, Internet resources and access are sufficiently important to many corporations and other employers that those employers link their office computer networks to the Internet and provide employees with direct or modem access to the office network (and thus to the Internet). Such access might be used by, for example, a corporation involved in scientific or medical research or manufacturing to enable corporate employees to exchange information and ideas with academic researchers in their fields.

15. Those who lack access to the Internet through their schools or employers still have a variety of ways they can access the Internet. Many communities across the country have established "free-nets" or community networks to provide their citizens with a local link to the Internet (and to provide local-oriented content and discussion groups). The first such community network, the Cleveland Free-Net Community Computer System, was established in 1986, and free-nets now exist in scores of communities as diverse as Richmond, Virginia, Tallahassee, Florida, Seattle, Washington, and San Diego, California. Individuals typically can access free-nets at little or no cost via modem connection or by using computers available in community buildings. Free-nets are often operated by a local library, educational institution, or non-profit community group.

16. Individuals can also access the Internet through many local libraries. Libraries often offer patrons use of computers that are linked to the Internet. In addition, some libraries offer telephone modem access to the libraries' computers, which are themselves connected to the Internet. Increasingly, patrons now use library services and resources without ever physically entering the library itself. Libraries typically provide such direct or modem access at no cost to the individual user.

17. Individuals can also access the Internet by patronizing an increasing number of storefront "computer coffee shops," where customers while they drink their coffee can use computers provided by the shop to access the Internet. Such Internet access is typically provided by the shop for a small hourly fee.

18. Individuals can also access the Internet through commercial and non-commercial "Internet service providers" that typically offer modem telephone access to a computer or computer network linked to the Internet. Many such providers including the members of plaintiff Commercial Internet Exchange Association are commercial entities offering Internet access for a monthly or hourly fee. Some Internet service providers, however, are non-profit organizations that offer free or very low cost access to the Internet. For example, the International Internet Association offers free modem access to the Internet upon request. Also, a number of trade or other non-profit associations offer Internet access as a service to members.

19. Another common way for individuals to access the Internet is through one of the major national commercial "online services" such as America Online, CompuServe, the Microsoft Network, or Prodigy. These online services offer nationwide computer networks (so that subscribers can dial-in to a local telephone number), and the services provide extensive and well organized content within their own proprietary computer networks. In addition to allowing access to the extensive content available within each online service, the services also allow subscribers to link to the much larger resources of the Internet. Full access to the online service (including access to the Internet) can be obtained for modest monthly or hourly fees. The major commercial online services have almost twelve million individual subscribers across the United States.

20. In addition to using the national commercial online services, individuals can also access the Internet using some (but not all) of the thousands of local dial-in computer services, often called "bulletin board systems" or "BBSs." With an investment of as little as $2,000.00 and the cost of a telephone line, individuals, non-profit organizations, advocacy groups, and businesses can offer their own dial-in computer "bulletin board" service where friends, members, subscribers, or customers can exchange ideas and information. BBSs range from single computers with only one telephone line into the computer (allowing only one user at a time), to single computers with many telephone lines into the computer (allowing multiple simultaneous users), to multiple linked computers each servicing multiple dial-in telephone lines (allowing multiple simultaneous users). Some (but not all) of these BBS systems offer direct or indirect links to the Internet. Some BBS systems charge users a nominal fee for access, while many others are free to the individual users.

21. Although commercial access to the Internet is growing rapidly, many users of the Internet such as college students and staff do not individually pay for access (except to the extent, for example, that the cost of computer services is a component of college tuition). These and other Internet users can access the Internet without paying for such access with a credit card or other form of payment.

Methods to Communicate Over the Internet

22. Once one has access to the Internet, there are a wide variety of different methods of communication and information exchange over the network. These many methods of communication and information retrieval are constantly evolving and are therefore difficult to categorize concisely. The most common methods of communications on the Internet (as well as within the major online services) can be roughly grouped into six categories:


   (1) one-to-one messaging (such as "e-mail"),


   (2) one-to-many messaging (such as "list-serv"),


   (3) distributed message databases (such as "USENET newsgroups"),


   (4) real time communication (such as "Internet Relay Chat"),


   (5) real time remote computer utilization (such as "telnet"), and


   (6) remote information retrieval (such as "ftp," "gopher," and the "World Wide Web").

Most of these methods of communication can be used to transmit text, data, computer programs, sound, visual images (i.e., pictures), and moving video images.

23. One-to-one messaging. One method of communication on the Internet is via electronic mail, or "e-mail," comparable in principle to sending a first class letter. One can address and transmit a message to one or more other people. E-mail on the Internet is not routed through a central control point, and can take many and varying paths to the recipients. Unlike postal mail, simple e-mail generally is not "sealed" or secure, and can be accessed or viewed on intermediate computers between the sender and recipient (unless the message is encrypted).

24. One-to-many messaging. The Internet also contains automatic mailing list services (such as "listservs"), [also referred to by witnesses as "mail exploders"] that allow communications about particular subjects of interest to a group of people. For example, people can subscribe to a "listserv" mailing list on a particular topic of interest to them. The subscriber can submit messages on the topic to the listserv that are forwarded (via e-mail), either automatically or through a human moderator overseeing the listserv, to anyone who has subscribed to the mailing list. A recipient of such a message can reply to the message and have the reply also distributed to everyone on the mailing list. This service provides the capability to keep abreast of developments or events in a particular subject area. Most listserv-type mailing lists automatically forward all incoming messages to all mailing list subscribers. There are thousands of such mailing list services on the Internet, collectively with hundreds of thousands of subscribers. Users of "open" listservs typically can add or remove their names from the mailing list automatically, with no direct human involvement. Listservs may also be "closed," i.e., only allowing for one's acceptance into the listserv by a human moderator.

25. Distributed message databases. Similar in function to listservs but quite different in how communications are transmitted are distributed message databases such as "USENET newsgroups." User-sponsored newsgroups are among the most popular and widespread applications of Internet services, and cover all imaginable topics of interest to users. Like listservs, newsgroups are open discussions and exchanges on particular topics. Users, however, need not subscribe to the discussion mailing list in advance, but can instead access the database at any time. Some USENET newsgroups are "moderated" but most are open access. For the moderated newsgroups,[1] all messages to the newsgroup are forwarded to one person who can screen them for relevance to the topics under discussion. USENET newsgroups are disseminated using ad hoc, peer to peer connections between approximately 200,000 computers (called USENET "servers") around the world. For unmoderated newsgroups, when an individual user with access to a USENET server posts a message to a newsgroup, the message is automatically forwarded to all adjacent USENET servers that furnish access to the newsgroup, and it is then propagated to the servers adjacent to those servers, etc. The messages are temporarily stored on each receiving server, where they are available for review and response by individual users. The messages are automatically and periodically purged from each system after a time to make room for new messages. Responses to messages, like the original messages, are automatically distributed to all other computers receiving the newsgroup or forwarded to a moderator in the case of a moderated newsgroup. The dissemination of messages to USENET servers around the world is an automated process that does not require direct human intervention or review.

26. There are newsgroups on more than fifteen thousand different subjects. In 1994, approximately 70,000 messages were posted to newsgroups each day, and those messages were distributed to the approximately 190,000 computers or computer networks that participate in the USENET newsgroup system. Once the messages reach the approximately 190,000 receiving computers or computer networks, they are available to individual users of those computers or computer networks. Collectively, almost 100,000 new messages (or "articles") are posted to newsgroups each day.

27. Real time communication. In addition to transmitting messages that can be later read or accessed, individuals on the Internet can engage in an immediate dialog, in "real time", with other people on the Internet. In its simplest forms, "talk" allows one-to-one communications and "Internet Relay Chat" (or IRC) allows two or more to type messages to each other that almost immediately appear on the others' computer screens. IRC is analogous to a telephone party line, using a computer and keyboard rather than a telephone. With IRC, however, at any one time there are thousands of different party lines available, in which collectively tens of thousands of users are engaging in conversations on a huge range of subjects. Moreover, one can create a new party line to discuss a different topic at any time. Some IRC conversations are "moderated" or include "channel operators."

28. In addition, commercial online services such as America Online, CompuServe, the Microsoft Network, and Prodigy have their own "chat" systems allowing their members to converse.

29. Real time remote computer utilization. Another method to use information on the Internet is to access and control remote computers in "real time" using "telnet." For example, using telnet, a researcher at a university would be able to use the computing power of a supercomputer located at a different university. A student can use telnet to connect to a remote library to access the library's online card catalog program.

30. Remote information retrieval. The final major category of communication may be the most well known use of the Internet the search for and retrieval of information located on remote computers. There are three primary methods to locate and retrieve information on the Internet.

31. A simple method uses "ftp" (or file transfer protocol) to list the names of computer files available on a remote computer, and to transfer one or more of those files to an individual's local computer.

32. Another approach uses a program and format named "gopher" to guide an individual's search through the resources available on a remote computer.

The World Wide Web

33. A third approach, and fast becoming the most well-known on the Internet, is the "World Wide Web." The Web utilizes a "hypertext" formatting language called hypertext markup language (HTML), and programs that "browse" the Web can display HTML documents containing text, images, sound, animation and moving video. Any HTML document can include links to other types of information or resources, so that while viewing an HTML document that, for example, describes resources available on the Internet, one can "click" using a computer mouse on the description of the resource and be immediately connected to the resource itself. Such "hyperlinks" allow information to be accessed and organized in very flexible ways, and allow people to locate and efficiently view related information even if the information is stored on numerous computers all around the world.

34. Purpose. The World Wide Web (W3C) was created to serve as the platform for a global, online store of knowledge, containing information from a diversity of sources and accessible to Internet users around the world. Though information on the Web is contained in individual computers, the fact that each of these computers is connected to the Internet through W3C protocols allows all of the information to become part of a single body of knowledge. It is currently the most advanced information system developed on the Internet, and embraces within its data model most information in previous networked information systems such as ftp, gopher, wais, and Usenet.

35. History. W3C was originally developed at CERN, the European Particle Physics Laboratory, and was initially used to allow information sharing within internationally dispersed teams of researchers and engineers. Originally aimed at the High Energy Physics community, it has spread to other areas and attracted much interest in user support, resource recovery, and many other areas which depend on collaborative and information sharing. The Web has extended beyond the scientific and academic community to include communications by individuals, non-profit organizations, and businesses.

36. Basic Operation. The World Wide Web is a series of documents stored in different computers all over the Internet. Documents contain information stored in a variety of formats, including text, still images, sounds, and video. An essential element of the Web is that any document has an address (rather like a telephone number). Most Web documents contain "links." These are short sections of text or image which refer to another document. Typically the linked text is blue or underlined when displayed, and when selected by the user, the referenced document is automatically displayed, wherever in the world it actually is stored. Links for example are used to lead from overview documents to more detailed documents, from tables of contents to particular pages, but also as cross-references, footnotes, and new forms of information structure.

37. Many organizations now have "home pages" on the Web. These are documents which provide a set of links designed to represent the organization, and through links from the home page, guide the user directly or indirectly to information about or relevant to that organization.

38. As an example of the use of links, if these Findings were to be put on a World Wide Web site, its home page might contain links such as those:

  • THE NATURE OF CYBERSPACE
  • CREATION OF THE INTERNET AND THE DEVELOPMENT OF CYBERSPACE
  • HOW PEOPLE ACCESS THE INTERNET
  • METHODS TO COMMUNICATE OVER THE INTERNET

39. Each of these links takes the user of the site from the beginning of the Findings to the appropriate section within this Adjudication. Links may also take the user from the original Web site to another Web site on another computer connected to the Internet. These links from one computer to another, from one document to another across the Internet, are what unify the Web into a single body of knowledge, and what makes the Web unique. The Web was designed with a maximum target time to follow a link of one tenth of a second.

40. Publishing. The World Wide Web exists fundamentally as a platform through which people and organizations can communicate through shared information. When information is made available, it is said to be "published" on the Web. Publishing on the Web simply requires that the "publisher" has a computer connected to the Internet and that the computer is running W3C server software. The computer can be as simple as a small personal computer costing less than $1500 dollars or as complex as a multi-million dollar mainframe computer. Many Web publishers choose instead to lease disk storage space from someone else who has the necessary computer facilities, eliminating the need for actually owning any equipment oneself.

41. The Web, as a universe of network accessible information, contains a variety of documents prepared with quite varying degrees of care, from the hastily typed idea, to the professionally executed corporate profile. The power of the Web stems from the ability of a link to point to any document, regardless of its status or physical location.

42. Information to be published on the Web must also be formatted according to the rules of the Web standards. These standardized formats assure that all Web users who want to read the material will be able to view it. Web standards are sophisticated and flexible enough that they have grown to meet the publishing needs of many large corporations, banks, brokerage houses, newspapers and magazines which now publish "online" editions of their material, as well as government agencies, and even courts, which use the Web to disseminate information to the public. At the same time, Web publishing is simple enough that thousands of individual users and small community organizations are using the Web to publish their own personal "home pages," the equivalent of individualized newsletters about that person or organization, which are available to everyone on the Web.

43. Web publishers have a choice to make their Web sites open to the general pool of all Internet users, or close them, thus making the information accessible only to those with advance authorization. Many publishers choose to keep their sites open to all in order to give their information the widest potential audience. In the event that the publishers choose to maintain restrictions on access, this may be accomplished by assigning specific user names and passwords as a prerequisite to access to the site. Or, in the case of Web sites maintained for internal use of one organization, access will only be allowed from other computers within that organization's local network.[2]

44. Searching the Web. A variety of systems have developed that allow users of the Web to search particular information among all of the public sites that are part of the Web. Services such as Yahoo, Magellan, Altavista, Webcrawler, and Lycos are all services known as "search engines" which allow users to search for Web sites that contain certain categories of information, or to search for key words. For example, a Web user looking for the text of Supreme Court opinions would type the words "Supreme Court" into a search engine, and then be presented with a list of World Wide Web sites that contain Supreme Court information. This list would actually be a series of links to those sites. Having searched out a number of sites that might contain the desired information, the user would then follow individual links, browsing through the information on each site, until the desired material is found. For many content providers on the Web, the ability to be found by these search engines is very important.

45. Common standards. The Web links together disparate information on an evergrowing number of Internet-linked computers by setting common information storage formats (HTML) and a common language for the exchange of Web documents (HTTP). Although the information itself may be in many different formats, and stored on computers which are not otherwise compatible, the basic Web standards provide a basic set of standards which allow communication and exchange of information. Despite the fact that many types of computers are used on the Web, and the fact that many of these machines are otherwise incompatible, those who "publish" information on the Web are able to communicate with those who seek to access information with little difficulty because of these basic technical standards.

46. A distributed system with no centralized control. Running on tens of thousands of individual computers on the Internet, the Web is what is known as a distributed system. The Web was designed so that organizations with computers containing information can become part of the Web simply by attaching their computers to the Internet and running appropriate World Wide Web software. No single organization controls any membership in the Web, nor is there any single centralized point from which individual Web sites or services can be blocked from the Web. From a user's perspective, it may appear to be a single, integrated system, but in reality it has no centralized control point.

47. Contrast to closed databases. The Web's open, distributed, decentralized nature stands in sharp contrast to most information systems that have come before it. Private information services such as Westlaw, Lexis/Nexis, and Dialog, have contained large storehouses of knowledge, and can be accessed from the Internet with the appropriate passwords and access software. However, these databases are not linked together into a single whole, as is the World Wide Web.

48. Success of the Web in research, education, and political activities. The World Wide Web has become so popular because of its open, distributed, and easy-to-use nature. Rather than requiring those who seek information to purchase new software or hardware, and to learn a new kind of system for each new database of information they seek to access, the Web environment makes it easy for users to jump from one set of information to another. By the same token, the open nature of the Web makes it easy for publishers to reach their intended audiences without having to know in advance what kind of computer each potential reader has, and what kind of software they will be using.

Restricting Access to Unwanted On-Line Material[3]

PICS

49. With the rapid growth of the Internet, the increasing popularity of the Web, and the existence of material online that some parents may consider inappropriate for their children, various entities have begun to build systems intended to enable parents to control the material which comes into their homes and may be accessible to their children. The World Wide Web Consortium launched the PICS ("Platform for Internet Content Selection") program in order to develop technical standards that would support parents' ability to filter and screen material that their children see on the Web.

50. The Consortium intends that PICS will provide the ability for third parties, as well as individual content providers, to rate content on the Internet in a variety of ways. When fully implemented, PICS-compatible World Wide Web browsers, Usenet News Group readers, and other Internet applications, will provide parents the ability to choose from a variety of rating services, or a combination of services.

51. PICS working group [PICS-WG] participants include many of the major online services providers, commercial internet access providers, hardware and software companies, major internet content providers, and consumer organizations. Among active participants in the PICS effort are:


   Adobe Systems, Inc.


   Apple Computer


   America Online


   AT & T


   Center for Democracy and Technology


   CompuServe


   Delphi Internet Services


   Digital Equipment Corporation


   IBM


   First floor


   First Virtual Holdings Incorporated


   France Telecom


   FTP Software


   Industrial Technology Research Institute of Taiwan


   Information Technology Association of America


   Institut National de Recherche en Informatique et en Automatique (INRIA)


   Interactive Services Association


   MCI


   Microsoft


   MIT/LCS/World Wide Web Consortium


   NCD


   NEC


   Netscape Communications Corporation


   NewView


   O'Reilly and Associates


   Open Market


   Prodigy Services Company


   Progressive Networks


   Providence Systems/Parental Guidance


   Recreational Software Advisory Council


   SafeSurf


   SoftQuad, Inc.


   Songline Studios


   Spyglass


   SurfWatch Software


   Telequip Corp.


   Time Warner Pathfinder


   Viacom Nickelodeon[4]

52. Membership in the PICS-WG includes a broad cross-section of companies from the computer, communications, and content industries, as well as trade associations and public interest groups. PICS technical specifications have been agreed to, allowing the Internet community to begin to deploy products and services based on the PICS-standards.

53. Until a majority of sites on the Internet have been rated by a PICS rating service, PICS will initially function as a "positive" ratings system in which only those sites that have been rated will be displayed using PICS compatible software. In other words, PICS will initially function as a site inclusion list rather than a site exclusion list. The default configuration for a PICS compatible Internet application will be to block access to all sites which have not been rated by a PICS rating service, while allowing access to sites which have a PICS rating for appropriate content.

Software

54. For over a year, various companies have marketed stand alone software that is intended to enable parents and other adults to limit the Internet access of children. Examples of such software include: Cyber Patrol, CYBERsitter, The Internet Filter, Net Nanny, Parental Guidance, SurfWatch, Netscape Proxy Server, and WebTrack. The market for this type of software is growing, and there is increasing competition among software providers to provide products.

Cyber Patrol

55. As more people, particularly children, began to use the Internet, Microsystems Software, Inc. decided to develop and market Internet software intended to empower parents to exercise individual choice over what material their children could access. Microsystems' stated intent is to develop a product which would give parents comfort that their children can reap the benefits of the Internet while shielding them from objectionable or otherwise inappropriate materials based on the parents' own particular tastes and values. Microsystems' product, Cyber Patrol, was developed to address this need.

56. Cyber Patrol was first introduced in August 1995, and is currently available in Windows and Macintosh versions. Cyber Patrol works with both direct Internet Access providers (ISPs, e.g., Netcom, PSI, UUnet), and Commercial Online Service Providers (e.g., America Online, CompuServ, Prodigy, Microsoft). Cyber Patrol is also compatible with all major World Wide Web browsers on the market (e.g., Netscape, Navigator, Mosaic, Prodigy's Legacy and Skimmer browsers, America Online, Netcom's NetCruiser, etc.). Cyber Patrol was the first parental empowerment application to be compatible with the PICS standard. In February of 1996, Microsystems put the first PICS ratings server on the Internet.

57. The CyberNOT list contains approximately 7000 sites in twelve categories. The software is designed to enable parents to selectively block access to any or all of the twelve CyberNOT categories simply by checking boxes in the Cyber Patrol Headquarters (the Cyber Patrol program manager). These categories are:


   Violence/Profanity: Extreme cruelty, physical or emotional acts against any animal or person which are primarily intended to hurt or inflict pain. Obscene words, phrases, and profanity defined as text that uses George Carlin's seven censored words more often than once every fifty messages or pages.


   Partial Nudity: Full or partial exposure of the human anatomy except when exposing genitalia.


   Nudity: Any exposure of the human genitalia.


   Sexual Acts (graphic or text): Pictures or text exposing anyone or anything involved in explicit sexual acts and lewd and lascivious behavior, including masturbation, copulation, pedophilia, intimacy and involving nude or partially nude people in heterosexual, bisexual, lesbian or homosexual encounters. Also includes phone sex ads, dating services, adult personals, CD-ROM and videos.


   Gross Depictions (graphic or text): Pictures or descriptive text of anyone or anything which are crudely vulgar, deficient in civility or behavior, or showing scatological impropriety. Includes such depictions as maiming, bloody figures, indecent depiction of bodily functions.


   Racism/Ethnic Impropriety: Prejudice or discrimination against any race or ethnic culture. Ethnic or racist jokes and slurs. Any text that elevates one race over another.


   Satanic/Cult: Worship of the devil; affinity for evil, wickedness. Sects or groups that potentially coerce individuals to grow, and keep, membership.


   Drugs/Drug Culture: Topics dealing with the use of illegal drugs for entertainment. This would exclude current illegal drugs used for medicinal purposes (e.g., drugs used to treat victims of AIDS). Includes substances used for other than their primary purpose to alter the individual's state of mind such as glue sniffing.


   Militant/Extremist: Extremely aggressive and combative behaviors, radicalism, advocacy of extreme political measures. Topics include extreme political groups that advocate violence as a means to achieve their goal.


   Gambling: Of or relating to lotteries, casinos, betting, numbers games, on-line sports or financial betting including non-monetary dares.


   Questionable/Illegal: Material or activities of a dubious nature which may be illegal in any or all jurisdictions, such as illegal business schemes, chain letters, software piracy, and copyright infringement.


   Alcohol, Beer & Wine: Material pertaining to the sale or consumption of alcoholic beverages. Also includes sites and information relating to tobacco products.

58. Microsystems employs people to search the Internet for sites containing material in these categories. Since new sites are constantly coming online, Microsystems updates the CyberNOT list on a weekly basis. Once installed on the home PC, the copy of Cyber Patrol receives automatic updates to the CyberNOT list over the Internet every seven days.

59. In February of 1996, Microsystems signed a licensing arrangement with CompuServe, one of the leading commercial online services with over 4.3 million subscribers. CompuServe provides Cyber Patrol free of charge to its subscribers. Microsystems the same month signed a licensing arrangement with Prodigy, another leading commercial online service with over 1.4 million subscribers. Prodigy will provide Cyber Patrol free of charge of its subscribers.

60. Cyber Patrol is also available directly from Microsystems for $49.95, which includes a six month subscription to the CyberNOT blocked sites list (updated automatically once every seven days). After six months, parents can receive six months of additional updates for $19.95, or twelve months for $29.95. Cyber Patrol Home Edition, a limited version of Cyber Patrol, is available free of charge on the Internet. To obtain either version, parents download a seven day demonstration version of the full Cyber Patrol product from the Microsystems Internet World Wide Web Server. At the end of the seven day trial period, users are offered the opportunity to purchase the complete version of Cyber Patrol or provide Microsystems some basic demographic information in exchange for unlimited use of the Home Edition. The demographic information is used for marketing and research purposes. Since January of 1996, over 10,000 demonstration copies of Cyber Patrol have been downloaded from Microsystems' Web site.

61. Cyber Patrol is also available from Retail outlets as NetBlocker Plus. NetBlocker Plus sells for $19.95, which includes five weeks of updates to the CyberNOT list.

62. Microsystems also sells Cyber Patrol into a growing market in schools. As more classrooms become connected to the Internet, many teachers want to ensure that their students can receive the benefit of the Internet without encountering material they deem educationally inappropriate.

63. Microsystems is working with the Recreational Software Advisory Council (RSAC), a non-profit corporation which developed rating systems for video games, to implement the RSAC rating system for the Internet.

64. The next release of Cyber Patrol, expected in second quarter of this year, will give parents the ability to use any PICS rating service, including the RSAC rating service, in addition to the Microsystems CyberNOT list.

65. In order to speed the implementation of PICS and encourage the development of PICS-compatible Internet applications, Microsystems maintains a server on the Internet which contains its CyberNOT list. The server provides software developers with access to a PICS rating service, and allows software developers to test their products' ability to interpret standard PICS labels. Microsystems is also offering its PICS client test program for Windows free of charge. The client program can be used by developers of PICS rating services to test their services and products.

SurfWatch

66. Another software product, SurfWatch, is also designed to allow parents and other concerned users to filter unwanted material on the Internet. SurfWatch is available for both Apple Macintosh, Microsoft Windows, and Microsoft Windows 95 Operating Systems, and works with direct Internet Access Providers (e.g., Netcom, PSI, UUnet, AT & T, and more than 1000 other Internet Service Providers).

67. The suggested retail price of Surf-Watch Software is $49.95, with a street price of between $20.00 and $25.00. The product is also available as part of CompuServe/Spry Inc.'s Internet in a Box for Kids, which includes access to Spry's Kids only Internet service and a copy of SurfWatch. Internet in a Box for Kids retails for approximately $30.00. The subscription service, which updates the SurfWatch blocked site list automatically with new sites each month, is available for $5.95 per month or $60.00 per year. The subscription is included as part of the Internet in a Box for Kids program, and is also provided as a low-cost option from Internet Service Providers.

68. SurfWatch is available at over 12,000 retail locations, including National stores such as Comp USA, Egghead Software, Computer City, and several national mail order outlets. SurfWatch can also be ordered directly from its own site on the World Wide Web, and through the Internet Shopping Network.

69. Plaintiffs America Online (AOL), Microsoft Network, and Prodigy all offer parental control options free of charge to their members. AOL has established an online area designed specifically for children. The "Kids Only" parental control feature allows parents to establish an AOL account for their children that accesses only the Kids Only channel on America Online.

70. AOL plans to incorporate PICS-compatible capability into its standard Web browser software, and to make available to subscribers other PICS-compatible Web browsers, such as the Netscape software.

71. Plaintiffs CompuServe and Prodigy give their subscribers the option of blocking all access to the Internet, or to particular media within their proprietary online content, such as bulletin boards and chat rooms.

72. Although parental control software currently can screen for certain suggestive words or for known sexually explicit sites, it cannot now screen for sexually explicit images unaccompanied by suggestive text unless those who configure the software are aware of the particular site.

73. Despite its limitations, currently available user-based software suggests that a reasonably effective method by which parents can prevent their children from accessing sexually explicit and other material which parents may believe is inappropriate for their children will soon be widely available.

Content on the Internet

74. The types of content now on the Internet defy easy classification. The entire card catalogue of the Carnegie Library is online, together with journals, journal abstracts, popular magazines, and titles of compact discs. The director of the Carnegie Library, Robert Croneberger, testified that on-line services are the emerging trend in libraries generally. Plaintiff Hotwired Ventures LLC organizes its Web site into information regarding travel, news and commentary, arts and entertainment, politics, and types of drinks. Plaintiff America Online, Inc., not only creates chat rooms for a broad variety of topics, but also allows members to create their own chat rooms to suit their own tastes. The ACLU uses an America Online chat room as an unmoderated forum for people to debate civil liberties issues. Plaintiffs' expert, Scott Bradner,[5] estimated that 15,000 newsgroups exist today, and he described his own interest in a newsgroup devoted solely to Formula 1 racing cars. America Online makes 15,000 bulletin boards available to its subscribers, who post between 200,000 and 250,000 messages each day. Another plaintiffs' expert, Howard Rheingold, participates in "virtual communities" that simulate social interaction. It is no exaggeration to conclude that the content on the Internet is as diverse as human thought.

75. The Internet is not exclusively, or even primarily, a means of commercial communication. Many commercial entities maintain Web sites to inform potential consumers about their goods and services, or to solicit purchases, but many other Web sites exist solely for the dissemination of non-commercial information. The other forms of Internet communication e-mail, bulletin boards, newsgroups, and chat rooms frequently have non-commercial goals. For the economic and technical reasons set forth in the following paragraphs, the Internet is an especially attractive means for not-for-profit entities or public interest groups to reach their desired audiences. There are examples in the parties' stipulation of some of the non-commercial uses that the Internet serves. Plaintiff Human Rights Watch, Inc., offers information on its Internet site regarding reported human rights abuses around the world. Plaintiff National Writers Union provides a forum for writers on issues of concern to them. Plaintiff Stop Prisoner Rape, Inc., posts text, graphics, and statistics regarding the incidence and prevention of rape in prisons. Plaintiff Critical Path AIDS Project, Inc., offers information on safer sex, the transmission of HIV, and the treatment of AIDS.

76. Such diversity of content on the Internet is possible because the Internet provides an easy and inexpensive way for a speaker to reach a large audience, potentially of millions. The start-up and operating costs entailed by communication on the Internet are significantly lower than those associated with use of other forms of mass communication, such as television, radio, newspapers, and magazines. This enables operation of their own Web sites not only by large companies, such as Microsoft and Time Warner, but also by small, not-for-profit groups, such as Stop Prisoner Rape and Critical Path AIDS Project. The Government's expert, Dr. Dan R. Olsen,[6] agreed that creation of a Web site would cost between $1,000 and $15,000, with monthly operating costs depending on one's goals and the Web site's traffic. Commercial online services such as America Online allow subscribers to create Web pages free of charge. Any Internet user can communicate by posting a message to one of the thousands of newsgroups and bulletin boards or by engaging in an on-line "chat", and thereby reach an audience worldwide that shares an interest in a particular topic.

77. The ease of communication through the Internet is facilitated by the use of hypertext markup language (HTML), which allows for the creation of "hyperlinks" or "links". HTML enables a user to jump from one source to other related sources by clicking on the link. A link might take the user from Web site to Web site, or to other files within a particular Web site. Similarly, by typing a request into a search engine, a user can retrieve many different sources of content related to the search that the creators of the engine have collected.

78. Because of the technology underlying the Internet, the statutory term "content provider,"[7] which is equivalent to the traditional "speaker," may actually be a hybrid of speakers. Through the use of HTML, for example, Critical Path and Stop Prisoner Rape link their Web sites to several related databases, and a user can immediately jump from the home pages of these organizations to the related databases simply by clicking on a link. America Online creates chat rooms for particular discussions but also allows subscribers to create their own chat rooms. Similarly, a newsgroup gathers postings on a particular topic and distributes them to the newsgroup's subscribers. Users of the Carnegie Library can read on-line versions of Vanity Fair and Playboy, and America Online's subscribers can peruse the New York Times, Boating, and other periodicals. Critical Path, Stop Prisoner Rape, America Online and the Carnegie Library all make available content of other speakers over whom they have little or no editorial control.

79. Because of the different forms of Internet communication, a user of the Internet may speak or listen interchangeably, blurring the distinction between "speakers" and "listeners" on the Internet. Chat rooms, e-mail, and newsgroups are interactive forms of communication, providing the user with the opportunity both to speak and to listen.

80. It follows that unlike traditional media, the barriers to entry as a speaker on the Internet do not differ significantly from the barriers to entry as a listener. Once one has entered cyberspace, one may engage in the dialogue that occurs there. In the argot of the medium, the receiver can and does become the content provider, and vice-versa.

81. The Internet is therefore a unique and wholly new medium of worldwide human communication.

Sexually Explicit Material on the Internet

82. The parties agree that sexually explicit material exists on the Internet. Such material includes text, pictures, and chat, and includes bulletin boards, newsgroups, and the other forms of Internet communication, and extends from the modestly titillating to the hardest-core.

83. There is no evidence that sexually-oriented material is the primary type of content on this new medium. Purveyors of such material take advantage of the same ease of access available to all users of the Internet, including establishment of a Web site.

84. Sexually explicit material is created, named, and posted in the same manner as material that is not sexually explicit. It is possible that a search engine can accidentally retrieve material of a sexual nature through an imprecise search, as demonstrated at the hearing. Imprecise searches may also retrieve irrelevant material that is not of a sexual nature. The accidental retrieval of sexually explicit material is one manifestation of the larger phenomenon of irrelevant search results.

85. Once a provider posts content on the Internet, it is available to all other Internet users worldwide. Similarly, once a user posts a message to a newsgroup or bulletin board, that message becomes available to all subscribers to that newsgroup or bulletin board. For example, when the UCR/California Museum of Photography posts to its Web site nudes by Edward Weston and Robert Mapplethorpe to announce that its new exhibit will travel to Baltimore and New York City, those images are available not only in Los Angeles, Baltimore, and New York City, but also in Cincinnati, Mobile, or Beijing wherever Internet users live. Similarly, the safer sex instructions that Critical Path posts to its Web site, written in street language so that the teenage receiver can understand them, are available not just in Philadelphia, but also in Provo and Prague. A chat room organized by the ACLU to discuss the United States Supreme Court's decision in FCC v. Pacifica Foundation would transmit George Carlin's seven dirty words to anyone who enters. Messages posted to a newsgroup dedicated to the Oklahoma City bombing travel to all subscribers to that newsgroup.

86. Once a provider posts its content on the Internet, it cannot prevent that content from entering any community. Unlike the newspaper, broadcast station, or cable system, Internet technology necessarily gives a speaker a potential worldwide audience. Because the Internet is a network of networks (as described above in Findings 1 through 4), any network connected to the Internet has the capacity to send and receive information to any other network. Hotwired Ventures, for example, cannot prevent its materials on mixology from entering communities that have no interest in that topic.

87. Demonstrations at the preliminary injunction hearings showed that it takes several steps to enter cyberspace. At the most fundamental level, a user must have access to a computer with the ability to reach the Internet (typically by way of a modem). A user must then direct the computer to connect with the access provider, enter a password, and enter the appropriate commands to find particular data. On the World Wide Web, a user must normally use a search engine or enter an appropriate address. Similarly, accessing newsgroups, bulletin boards, and chat rooms requires several steps.

88. Communications over the Internet do not "invade" an individual's home or appear on one's computer screen unbidden. Users seldom encounter content "by accident." A document's title or a description of the document will usually appear before the document itself takes the step needed to view it, and in many cases the user will receive detailed information about a site's content before he or she need take the step to access the document. Almost all sexually explicit images are preceded by warnings as to the content. Even the Government's witness, Agent Howard Schmidt, Director of the Air Force Office of Special Investigation, testified that the "odds are slim" that a user would come across a sexually explicit site by accident.

89. Evidence adduced at the hearing showed significant differences between Internet communications and communications received by radio or television. Although content on the Internet is just a few clicks of a mouse away from the user, the receipt of information on the Internet requires a series of affirmative steps more deliberate and directed than merely turning a dial. A child requires some sophistication and some ability to read to retrieve material and thereby to use the Internet unattended.

Obstacles to Age Verification on the Internet

90. There is no effective way to determine the identity or the age of a user who is accessing material through e-mail, mail exploders, newsgroups or chat rooms. An e-mail address provides no authoritative information about the addressee, who may use an e-mail "alias" or an anonymous remailer. There is also no universal or reliable listing of e-mail addresses and corresponding names or telephone numbers, and any such listing would be or rapidly become incomplete. For these reasons, there is no reliable way in many instances for a sender to know if the e-mail recipient is an adult or a minor. The difficulty of e-mail age verification is compounded for mail exploders such as listservs, which automatically send information to all e-mail addresses on a sender's list. Government expert Dr. Olsen agreed that no current technology could give a speaker assurance that only adults were listed in a particular mail exploder's mailing list.

91. Because of similar technological difficulties, individuals posting a message to a newsgroup or engaging in chat room discussions cannot ensure that all readers are adults, and Dr. Olsen agreed. Although some newsgroups are moderated, the moderator's control is limited to what is posted and the moderator cannot control who receives the messages.

92. The Government offered no evidence that there is a reliable way to ensure that recipients and participants in such fora can be screened for age. The Government presented no evidence demonstrating the feasibility of its suggestion that chat rooms, newsgroups and other fora that contain material deemed indecent could be effectively segregated to "adult" or "moderated" areas of cyberspace.

93. Even if it were technologically feasible to block minors' access to newsgroups and similar fora, there is no method by which the creators of newsgroups which contain discussions of art, politics or any other subject that could potentially elicit "indecent" contributions could limit the blocking of access by minors to such "indecent" material and still allow them access to the remaining content, even if the overwhelming majority of that content was not indecent.

94. Likewise, participants in MUDs (Multi-User Dungeons) and MUSEs (Multi-User Simulation Environments) do not know whether the other participants are adults or minors. Although MUDs and MUSEs require a password for permanent participants, they need not give their real name nor verify their age, and there is no current technology to enable the administrator of these fantasy worlds to know if the participant is an adult or a minor.

95. Unlike other forms of communication on the Internet, there is technology by which an operator of a World Wide Web server may interrogate a user of a Web site. An HTML document can include a fill-in-the-blank "form" to request information from a visitor to a Web site, and this information can be transmitted back to the Web server and be processed by a computer program, usually a Common Gateway Interface (cgi) script. The Web server could then grant or deny access to the information sought. The cgi script is the means by which a Web site can process a fill-in form and thereby screen visitors by requesting a credit card number or adult password.

96. Content providers who publish on the World Wide Web via one of the large commercial online services, such as America Online or CompuServe, could not use an online age verification system that requires cgi script because the server software of these online services available to subscribers cannot process cgi scripts. There is no method currently available for Web page publishers who lack access to cgi scripts to screen recipients online for age.

The Practicalities of the Proffered Defenses

Note: The Government contends the CDA makes available three potential defenses to all content providers on the Internet: credit card verification, adult verification by password or adult identification number, and "tagging".

Credit Card Verification

97. Verification[8] of a credit card number over the Internet is not now technically possible. Witnesses testified that neither Visa nor Mastercard considers the Internet to be sufficiently secure under the current technology to process transactions in that manner. Although users can and do purchase products over the Internet by transmitting their credit card number, the seller must then process the transaction with Visa or Mastercard offline using phone lines in the traditional way. There was testimony by several witnesses that Visa and Mastercard are in the process of developing means of credit card verification over the Internet.

98. Verification by credit card, if and when operational, will remain economically and practically unavailable for many of the non-commercial plaintiffs in these actions. The Government's expert "suspect[ed]" that verification agencies would decline to process a card unless it accompanied a commercial transaction. There was no evidence to the contrary.

99. There was evidence that the fee charged by verification agencies to process a card, whether for a purchase or not, will preclude use of the credit-card verification defense by many non-profit, non-commercial Web sites, and there was no evidence to the contrary. Plaintiffs' witness Patricia Nell Warren, an author whose free Web site allows users to purchase gay and lesbian literature, testified that she must pay $1 per verification to a verification agency. Her Web site can absorb this cost because it arises in connection with the sale of books available there.

100. Using credit card possession as a surrogate for age, and requiring use of a credit card to enter a site, would impose a significant economic cost on non-commercial entities. Critical Path, for example, received 3,300 hits daily from February 4 through March 4, 1996. If Critical Path must pay a fee every time a user initially enters its site, then, to provide free access to its non-commercial site, it would incur a monthly cost far beyond its modest resources. The ACLU's Barry Steinhardt testified that maintenance of a credit card verification system for all visitors to the ACLU's Web site would require it to shut down its Web site because the projected cost would exceed its budget.

101. Credit card verification would significantly delay the retrieval of information on the Internet. Dr. Olsen, the expert testifying for the Government, agreed that even "a minute is [an] absolutely unreasonable [delay] ... [P]eople will not put up with a minute." Plaintiffs' expert Donna Hoffman similarly testified that excessive delay disrupts the "flow" on the Internet and stifles both "hedonistic" and "goal-directed" browsing.

102. Imposition of a credit card requirement would completely bar adults who do not have a credit card and lack the resources to obtain one from accessing any blocked material. At this time, credit card verification is effectively unavailable to a substantial number of Internet content providers as a potential defense to the CDA.

Adult Verification by Password

103. The Government offered very limited evidence regarding the operation of existing age verification systems, and the evidence offered was not based on personal knowledge. AdultCheck and Verify, existing systems which appear to be used for accessing commercial pornographic sites, charge users for their services. Dr. Olsen admitted that his knowledge of these services was derived primarily from reading the advertisements on their Web pages. He had not interviewed any employees of these entities, had not personally used these systems, had no idea how many people are registered with them, and could not testify to the reliability of their attempt at age verification.

104. At least some, if not almost all, non-commercial organizations, such as the ACLU, Stop Prisoner Rape or Critical Path AIDS Project, regard charging listeners to access their speech as contrary to their goals of making their materials available to a wide audience free of charge.

105. It would not be feasible for many non-commercial organizations to design their own adult access code screening systems because the administrative burden of creating and maintaining a screening system and the ongoing costs involved is beyond their reach. There was testimony that the costs would be prohibitive even for a commercial entity such as HotWired, the online version of Wired magazine.

106. There is evidence suggesting that adult users, particularly casual Web browsers, would be discouraged from retrieving information that required use of a credit card or password. Andrew Anker testified that HotWired has received many complaints from its members about HotWired's registration system, which requires only that a member supply a name, e-mail address and self-created password. There is concern by commercial content providers that age verification requirements would decrease advertising and revenue because advertisers depend on a demonstration that the sites are widely available and frequently visited.

107. Even if credit card verification or adult password verification were implemented, the Government presented no testimony as to how such systems could ensure that the user of the password or credit card is in fact over 18. The burdens imposed by credit card verification and adult password verification systems make them effectively unavailable to a substantial number of Internet content providers.

The Government's "Tagging" Proposal

108. The feasibility and effectiveness of "tagging" to restrict children from accessing "indecent" speech, as proposed by the Government has not been established. "Tagging" would require content providers to label all of their "indecent" or "patently offensive" material by imbedding a string of characters, such as "XXX," in either the URL or HTML. If a user could install software on his or her computer to recognize the "XXX" tag, the user could screen out any content with that tag. Dr. Olsen proposed a "-L18" tag, an idea he developed for this hearing in response to Mr. Bradner's earlier testimony that certain tagging would not be feasible.

109. The parties appear to agree that it is technologically feasible "trivial", in the words of plaintiffs' expert to imbed tags in URLs and HTML, and the technology of tagging underlies both plaintiffs' PICS proposal and the Government's "-L18" proposal.

110. The Government's tagging proposal would require all content providers that post arguably "indecent" material to review all of their online content, a task that would be extremely burdensome for organizations that provide large amounts of material online which cannot afford to pay a large staff to review all of that material. The Carnegie Library would be required to hire numerous additional employees to review its online files at an extremely high cost to its limited budget. The cost and effort would be substantial for the Library and frequently prohibitive for others. Witness Kiroshi Kuromiya testified that it would be impossible for his organization, Critical Path, to review all of its material because it has only one full and one part-time employee.

111. The task of screening and tagging cannot be done simply by using software which screens for certain words, as Dr. Olsen acknowledged, and we find that determinations as to what is indecent require human judgment.

112. In lieu of reviewing each file individually, a content provider could tag its entire site but this would prevent minors from accessing much material that is not "indecent" under the CDA.

113. To be effective, a scheme such as the -L18 proposal would require a worldwide consensus among speakers to use the same tag to label "indecent" material. There is currently no such consensus, and no Internet speaker currently labels its speech with the -L18 code or with any other widely-recognized label.

114. Tagging also assumes the existence of software that recognizes the tags and takes appropriate action when it notes tagged speech. Neither commercial Web browsers nor user-based screening software is currently configured to block a -L18 code. Until such software exists, all speech on the Internet will continue to travel to whomever requests it, without hindrance. Labelling speech has no effect in itself on the transmission (or not) of that speech. Neither plaintiffs nor the Government suggest that tagging alone would shield minors from speech or insulate a speaker from criminal liability under the CDA. It follows that all speech on any topic that is available to adults will also be available to children using the Internet (unless it is blocked by screening software running on the computer the child is using).

115. There is no way that a speaker can use current technology to know if a listener is using screening software.

116. Tags can not currently activate or deactivate themselves depending on the age or location of the receiver. Critical Path, which posts on-line safer sex instructions, would be unable to imbed tags that block its speech only in communities where it may be regarded as indecent. Critical Path, for example, must choose either to tag its site (blocking its speech in all communities) or not to tag, blocking its speech in none.

The Problems of Offshore Content and Caching

117. A large percentage, perhaps 40% or more, of content on the Internet originates outside the United States. At the hearing, a witness demonstrated how an Internet user could access a Web site of London (which presumably is on a server in England), and then link to other sites of interest in England. A user can sometimes discern from a URL that content is coming from overseas, since InterNIC allows a content provider to imbed a country code in a domain name.[9] Foreign content is otherwise indistinguishable from domestic content (as long as it is in English), since foreign speech is created, named, and posted in the same manner as domestic speech. There is no requirement that foreign speech contain a country code in its URL. It is undisputed that some foreign speech that travels over the Internet is sexually explicit.

118. The use of "caching" makes it difficult to determine whether the material originated from foreign or domestic sources. Because of the high cost of using the trans-Atlantic and trans-Pacific cables, and because the high demand on those cables leads to bottleneck delays, content is often "cached", or temporarily stored, on servers in the United States. Material from a foreign source in Europe can travel over the trans-Atlantic cable to the receiver in the United States, and pass through a domestic caching server which then stores a copy for subsequent retrieval. This domestic caching server, rather than the original foreign server, will send the material from the cache to the subsequent receivers, without placing a demand on the trans-oceanic cables. This shortcut effectively eliminates most of the distance for both the request and the information and, hence, most of the delay. The caching server discards the stored information according to its configuration (e.g., after a certain time or as the demand for the information diminishes). Caching therefore advances core Internet values: the cheap and speedy retrieval of information.

119. Caching is not merely an international phenomenon. Domestic content providers store popular domestic material on their caching servers to avoid the delay of successive searches for the same material and to decrease the demand on their Internet connection. America Online can cache the home page of the New York Times on its servers when a subscriber first requests it, so that subsequent subscribers who make the same request will receive the same home page, but from America Online's caching service rather than from the New York Times's server.[10]

120. Put simply, to follow the example in the prior paragraph, America Online has no control over the content that the New York Times posts to its Web site, and the New York Times has no control over America Online's distribution of that content from a caching server.

Anonymity

121. Anonymity is important to Internet users who seek to access sensitive information, such as users of the Critical Path AIDS Project's Web site, the users, particularly gay youth, of Queer Resources Directory, and users of Stop Prisoner Rape (SPR). Many members of SPR's mailing list have asked to remain anonymous due to the stigma of prisoner rape.

Plaintiffs' Choices under the CDA

122. Many speakers who display arguably indecent content on the Internet must choose between silence and the risk of prosecution. The CDA's defenses credit card verification, adult access codes, and adult personal identification numbers are effectively unavailable for non-commercial, not-for-profit entities.

123. The plaintiffs in this action are businesses, libraries, non-commercial and not-for-profit organizations, and educational societies and consortia. Although some of the material that plaintiffs post online such as information regarding protection from AIDS, birth control or prison rape is sexually explicit and may be considered "indecent" or "patently offensive" in some communities, none of the plaintiffs is a commercial purveyor of what is commonly termed "pornography."

FINDINGS OF FACT (Shea)

We enter the following findings of fact, many of which are undisputed, the subject of stipulations by the parties, or submitted by the defendant and adopted by us, pursuant to Rule 52(a) of the Federal Rules of Civil Procedure. Although we here consider a so-called facial challenge to a statute, we deemed it appropriate and necessary in the unusual circumstances presented here, and a reasonable exercise of our discretion, to establish a basic record of the facts regarding the new and evolving communications media that is the subject of this legislation.

Section 223(d) targets the use of an “interactive computer service” to send or display patently offensive materials. Although § 223 itself contains no definition of that term, the definition applicable to the new 47 U.S.C. § 230—also added by the CDA—makes clear that the term encompasses means of making “content”[11] available to multiple users both on the vast web of linked networks popularly known as “the Internet” and on other information systems (such as electronic bulletin boards maintained by educational institutions or nonprofit organizations) not physically linked to the Internet. See Pub.L. No. 104–104, § 509(e)(2), 110 Stat. at 139 (to be codified at 47 U.S.C. § 230(e)(2)). We draw upon the stipulations of the parties and the testimony adduced at the three-day evidentiary hearing to describe: (1) the nature of the medium targeted by § 223(d), focusing in part on the degree of control that those who transmit content have over who will receive it; (2) the availability of certain categories of potentially objectionable material on line; (3) the development of software and labeling standards enabling parents to limit their children's exposure to objectionable on-line content; and (4) the potential for tagging and verification procedures that content providers can use in an effort to shield minors from sexually explicit content that they provide.[12] As we do so, we unavoidably—and with apologies to all others with a similar aversion to “cyberspeak”—adopt some of the terminology that has developed in conjunction with this technology. We endeavor, to the extent possible, to avoid the jargon of this field, and to define our terms wherever possible, for the sake of the clarity of this record and this opinion, as well as for the benefit of any reader required to review our work.

A. The Development of the Internet

Although “the Internet” now formally describes a collection of more than 50,000 networks linking some nine million host computers in ninety countries, it has existed for nearly three decades on a much smaller scale. What we now refer to as the Internet grew out of an experimental project of the Department of Defense's Advanced Research Projects Administration (“ARPA”) designed to provide researchers with direct access to supercomputers at a few key laboratories and to facilitate the reliable transmission of vital communications. (Declaration of William J. Hoffman (“Hoffman Decl.”), Ex. 4, at 11–12) ARPA supplied funds to link computers operated by the military, defense contractors, and universities conducting defense-related research through dedicated phone lines, creating a “network” known as ARPANet. (Parties' Stipulations in Preparation for Preliminary Injunction Hearing (“Joint Stip.”) ¶¶ 6–7; Hoffman Decl., Ex. 3, at 3; id. Ex. 4, at 11) Programs on the linked computers implemented a technical scheme known as “packet-switching,” through which a message from one computer to another would be subdivided into smaller, separately addressed pieces of data, known as “packets,” sent independently to the message's destination and reassembled upon arrival. (Joint Stip. ¶ 9) Each computer on the network was in turn linked to several other computers, creating any number of routes that a communication from one computer could follow to reach its destination. If part of the network were damaged, a portion of the message could be re-routed automatically over any other path to its ultimate destination, a characteristic of the network intended initially to preserve its operability in the event of enemy attack. (Id. ¶¶ 7–8; Hoffman Decl., Ex. 3, at 3; id. Ex. 4, at 12)

Having successfully implemented a system for the reliable transfer of information over a computer network, ARPA began to support the development of communications protocols for transferring data between different types of computer networks. Universities, research facilities, and commercial entities began to develop and link together their own networks implementing these protocols; these networks included a high-speed “backbone” network known as NSFNet, sponsored by the National Science Foundation, smaller regional networks, and, eventually, large commercial networks run by organizations such as Sprint, IBM, and Performance Systems International (commonly known as “PSI”). (Hoffman Decl., Ex. 3, at 3; id. Ex. 4, at 13–14) As faster networks developed, most network traffic shifted away from ARPANet, which formally ceased operations in 1990. (Id. Ex. 3, at 3) What we know as “the Internet” today is the series of linked, overlapping networks that gradually supplanted ARPANet. Because the Internet links together independent networks that merely use the same data transfer protocols, it cannot be said that any single entity or group of entities controls, or can control, the content made publicly available on the Internet or limits, or can limit, the ability of others to access public content. Rather, the resources available to one with Internet access are located on individual computers around the world. (Joint Stip. ¶ 11)

It is estimated that as many as forty million individuals have access to the information and tools of the Internet, and that figure is expected to grow to 200 million by the year 1999. (Id. ¶ 3) Access to the Internet can take any one of several forms. First, many educational institutions, businesses, libraries, and individual communities maintain a computer network linked directly to the Internet and issue account numbers and passwords enabling users to gain access to the network directly or by modem.[13] (Id. ¶¶ 12–14) Second, “Internet service providers,” generally commercial entities charging a monthly fee, offer modem access to computers or networks linked directly to the Internet. (Id. ¶ 16) Third, national commercial “on-line services”—such as America Online, CompuServe, Prodigy, and Microsoft Network—allow subscribers to gain access to the Internet while providing extensive content within their own proprietary networks. (Id. ¶ 17) Finally, organizations and businesses can offer access to electronic bulletin-board systems—which, like national on-line services, provide certain proprietary content; some bulletin-board systems in turn offer users links to the Internet. (Id. ¶ 18)

B. Categories of Internet Use

For our purposes, there are two loose and overlapping categories of Internet use. First, an individual who has secured access to the Internet can correspond or exchange views with one or many other Internet users. Second, a user can locate and retrieve information available on other computers. We explore these categories in greater detail below. As will become clear, distinctions in how Internet content is transmitted affect the degree of control that providers of content have over who will be able to gain access to their communications;[14] we will return to the legal significance of these distinctions at a later juncture. For any communication to take place over the Internet, two pieces of software,[15] adhering to the same communications protocol, are required. A user must have access to certain kinds of “client” software, which enables his computer to communicate with and make requests of remote computers where information is stored; these remote computers must be running “server” software, which provides information in response to requests by client software. (Declaration of Dr. Dan R. Olsen, Jr. (“Olsen Decl.”), ¶¶ 13–14)

1. Communicating with Other Internet Users

Perhaps the most widely used Internet service is electronic mail, or “e-mail.” Using any one of dozens of available “mailers”—client software capable of reading and writing e-mail—a user is able to address and transmit a message to one or more specific individuals. (Joint Stip. ¶ 21) A user can also “subscribe” to an electronic mailing list on a topic of interest; the user receives a copy of messages posted by other subscribers and, in turn, can post messages for forwarding to the full mailing list. Once a mailing list is established, it is typically maintained using a “mail exploder”—a program such as “listserv” running on the server on which the list resides—that automatically (i.e., without human intervention) responds to a user's request to be added to or removed from the list of subscribers and retransmits messages posted by a subscriber to others on the mailing list. (Id. ¶ 22) Some mailing lists are “closed”: a user's request to join the list requires the approval of an individual who maintains the list. (Id.) Mailing lists (both open and closed) may also be “moderated”: all messages posted to the list are forwarded to a moderator, who approves certain messages and retransmits them to subscribers. (Id.) An individual sending a message that will be retransmitted by a mail exploder program has no way of knowing the e-mail addresses of other subscribers. (Olsen Decl. ¶ 19; Testimony of Gordon C. Galligher, Jr., Tr. at 181) Even if the user could obtain an e-mail address for each subscriber to a particular list, those addresses alone would provide no authoritative information about subscribers. There is no directory that identifies persons using a certain e-mail address. In addition, a user can avoid disclosing his true e-mail address by developing an e-mail “alias” or by using an “anonymous remailer”—a server that purges identifying information from a communication before forwarding it to its destination. (Defendant's Response to Plaintiff's Request for Admissions (“Defendant's Adm.”) No. 22; Galligher Test., Tr. at 173)

Internet users may also transmit or receive “articles” posted daily to thousands of discussion groups, arranged by subject matter and known as “newsgroups,” available through an electronic bulletin-board system known as “Usenet.” When a user with access to a Usenet server—that is, a computer participating in the Usenet system—posts an article to a particular newsgroup, the server automatically forwards the article to adjacent Usenet servers, which in turn forward it to other servers, until the article is available on all Usenet sites that furnish access to the newsgroup in question. (Joint Stip. ¶ 23) Once a message reaches a particular Usenet site, it is temporarily stored there so that individual users—running client software, known as a “newsreader,” capable of sorting articles according to header information identifying the newsgroup to which the article was posted—can review and respond to the message. (Id.; Hoffman Decl., Ex. 4, at 129) Some Usenet newsgroups are moderated; messages to the newsgroup are forwarded to an individual who selects those appropriate for distribution. (Joint Stip. ¶ 23) Because Usenet articles are distributed to (and made available on) multiple servers, one who posts an article to a newsgroup has no way of knowing who will choose to retrieve it, whether or not the newsgroup is moderated. (Galligher Test., Tr. at 170, 174–75) There is no newsgroup equivalent of a “closed” mailing list: access to a particular newsgroup can only be limited by restricting the number of servers participating in the newsgroup. (Testimony of Clay Shirky, Tr. at 251)

The Internet also offers opportunities for multiple users to interact in real time. Using a program called “Talk,” two users can exchange messages while they are both on line; a message typed on one user's computer will appear almost immediately on the other's screen. (Joint Stip. ¶ 25) Servers running so-called “chat” software, such as Internet Relay Chat (“IRC”), permit multiple users to converse by selecting one of many discussion “channels” active at any time. Commercial on-line services such as America Online, CompuServe, Prodigy, and the Microsoft Network offer their own chat systems for their members. (Id. ¶ 26) Having joined a channel, the user can see and read messages transmitted by other users, each identified by a name the user selects upon joining the channel. (Id. ¶ 25) Individual participants in IRC discussions know other participants only by the names they choose upon entering the discussion; users can participate anonymously by using a pseudonym.

2. Locating and Retrieving Information on the Internet

Individuals with Internet access can take advantage of a number of tools for locating and retrieving information and resources stored on remote computers. One who wishes to make certain articles, files, or software available to other users will set up a server, adhering to certain communications protocols, capable of retrieving and presenting stored information in response to a request from client software using the same communications protocol. (Olsen Decl. ¶¶ 13, 16; Galligher Test., Tr. at 131)

a. File–Transfer Protocol (“FTP”)

One type of software implements a set of conventions for copying files from a host computer known as “file-transfer protocol” (“FTP”). With appropriate client software, a user with an account on the host computer can contact the server, view a directory of available files, and copy one or more of those files to his own computer. In addition to making files available to users with accounts, thousands of content providers also make files available for “anonymous” retrieval by users who do not possess an account on the host computer.[16] A content provider who makes files available for retrieval by anonymous FTP has no way of discerning who gains access to the files.

b. “Gopher ” Servers

A second type of server software capable of making available the resources of a host computer is known as a “gopher” program. (Joint Stip. ¶ 30, Hoffman Decl., Ex. 3, at 5) A gopher server presents information in a set of menus, enabling a user who gains access to the server to select a series of increasingly narrow menu items before locating a desired file that can be displayed on or copied to the user's computer.[17] (Galligher Test., Tr. at 122; Hoffman Decl., Ex. 3, at 5) A content provider who maintains a gopher server ordinarily has no way of knowing who will gain access to the information made available.

c. The World Wide Web

The third and perhaps best known method of locating and accessing information on the Internet is by exploring the World Wide Web. Documents available on the Web are not collected in any central location; rather, they are stored on servers around the world running Web server software. (Joint Stip. ¶¶ 31, 38, 40) To gain access to the content available on the Web, a user must have a Web “browser”—client software, such as Netscape Navigator, Mosaic, or Internet Explorer, capable of displaying documents formatted in “hypertext markup language” (“HTML”), the standard Web formatting language. (Galligher Test., Tr. at 125; Joint Stip. ¶¶ 31, 43) Each document has an address, known as a Uniform Resource Locator (“URL”), identifying, among other things, the server on which it resides; most documents also contain “links”—highlighted text or images that, when selected by the user, permit him to view another, related Web document. (Joint Stip. ¶ 34) Because Web servers are linked to the Internet through a common communications protocol, known as hypertext transfer protocol (“HTTP”), a user can move seamlessly between documents, regardless of their location; when a user viewing a document located on one server selects a link to a document located elsewhere, the browser will automatically contact the second server and display the document. (Joint Stip. ¶¶ 34, 37) Some types of Web client software also permit users to gain access to resources available on FTP and gopher sites.

A number of “search engines”—such as Yahoo, Magellan, Alta Vista, WebCrawler, and Lycos—are available to help users navigate the World Wide Web.[18] For example, the service Yahoo maintains a directory of documents available on various Web servers. A user can gain access to Yahoo's server and type a string of characters as a search request. Yahoo returns a list of documents whose entries in the Yahoo directory match the search string and organizes the list of documents by category. (Galligher Test., Tr. at 134; Plaintiff's Ex. 3) Search engines make use of software capable of automatically contacting various Web sites and extracting relevant information. Some search engines, such as Alta Vista, store the information in a database and return it in response to a user request. Others, such as Yahoo, employ a group of individuals to determine whether and how a site should be categorized in the Yahoo directory. (Galligher Test., Tr. at 137; Supplemental Declaration of William J. Hoffman (“Hoffman Supp.Decl.”) Ex. A, at 39–42 (Testimony of Donna L. Hoffman in ACLU/ALA ))

As the growth in Internet use and the wide availability of tools and resources to those with access to the Internet suggest, the Internet presents extremely low entry barriers to those who wish to convey Internet content or gain access to it. In particular, a user wishing to communicate through e-mail, newsgroups, or Internet Relay Chat need only have access to a computer with appropriate software and a connection to the Internet, usually available for a low monthly fee. The user then in a sense becomes a public “speaker,” able to convey content, at relatively low cost, to users around the world to whom it may be of interest. Those who possess more sophisticated equipment and greater technical expertise can make content available on the Internet for retrieval by others (known or unknown) by running a server supporting anonymous FTP, a gopher server, or a Web server. Yet content providers need not necessarily run their own servers or have the programming expertise to construct their own sites; they can lease space on a Web server from another or create a “home page” through an on-line commercial service.

The ease of entry of many speakers sets interactive computer systems apart from any other more traditional communications medium that Congress has attempted to regulate in the past. With one-way media such as radio and television broadcasting or cable programming, a user is merely a listener or viewer; in the CDA, Congress sought to target “interactive” computer systems through which a listener or viewer, by definition, has the power to become a speaker. The relative ease of speaker entry and the relative parity among speakers accounts for the unprecedented and virtually unlimited opportunities for political discourse, cultural development, and intellectual activity that Congress found to characterize emerging communication technologies.

In seeking to describe the range of tools and opportunities for Internet users to “speak,” we recognize that the categories we delineate are far from clean and the technology is far from static. Indeed, by all indications, the way that we conceptualize various media that we have traditionally viewed as distinct—such as cable television, telephones, and computer networks—will change dramatically as these media “converge” into common forms of communication. See Denver Area Educ. Telecommunications Consortium v. FCC (“Denver Area Consortium”), 518 U.S. 727, –––– & n. 4, 116 S.Ct. 2374, 2402 & n. 4, 135 L.Ed.2d 888 (1996) (Souter, J., concurring); see also Jerry Berman & Daniel J. Weitzner, Abundance and User Control: Renewing the Democratic Heart of the First Amendment in the Age of Interactive Media, 104 Yale L.J. 1619, 1619 n. 1 (1995); Art Kramer, Netwatch: The AJC's Daily Online Guide, Atl.J. & Const., May 29, 1996, at B04 (describing cable modem technology designed to offer Internet access through existing cable television connections) (Hoffman Supp.Decl., Ex. C, at 3–4). Of course, our findings of fact are necessarily time-bound. We can only determine whether the statutory provision at issue here, in light of the technology available during the pendency of this case, comports with the First Amendment.

C. Sexually Explicit Content on the Internet

It is undisputed that there exists some content on the Internet that is—to use the Government's phrase—“sexually explicit.” (Defendant's Memorandum of Law, filed March 19, 1996, at 11) The term “sexually explicit” is descriptive rather than legal and does not appear in the statutory provision at issue, but the Government employs it as a shorthand to describe Internet content depicting “sexual or excretory activities or organs”—possibly though not necessarily in a patently offensive way. (Defendant's Supplemental Memorandum of Law (“Defendant's Supp. Memo.”), filed June 7, 1996, at 9) That is, the Government does not contend that all sexually explicit material is “patently offensive” and therefore within the scope of the CDA, but claims that there is certainly content available on the Internet that is both sexually explicit and patently offensive.

The testimony and demonstration of one of the Government's expert witnesses, Howard Schmidt, Director of the Air Force Office of Special Investigations, amply confirmed the availability of sexually explicit material on line. Nevertheless, there is no persuasive evidence in the record to suggest, much less prove, that sexually explicit material easily “assaults” an unknowing user—as in other media, most notably television and radio—or that any substantial proportion of Internet content is sexually explicit.

1. Ease of Access to Sexually Explicit Content

The Government urges us to conclude that an Internet user can easily stumble upon sexually explicit material. (Defendant's Post–Hearing Memorandum of Law (“Defendant's Post–Hearing Memo.”), filed May 28, 1996, at 31–32) It is important to begin with the general observation that, with the exception of e-mail, no content appears on a user's screen without the user having first taken some affirmative step. One wishing to read articles posted to a newsgroup must connect to a Usenet server and select the relevant group. To retrieve a file through anonymous FTP or access a gopher server, the user must search for or know the address of a particular server. To gain access to content on the World Wide Web, a user must know the URL of a relevant site or type a keyword into one of several available search engines.

Schmidt's demonstration focused mainly on the availability of sexually explicit content on the World Wide Web. In the absence of any screening software or filter, a user determined to view a site containing sexually explicit material can certainly do so, either by typing a known URL or by searching for key words. One sexually explicit site may, in turn, contain “links” to other such sites. (Defendant's Exs. 13, 16, 17, 26, 29, 32; Schmidt Test., Tr. at 401–02) While ordinarily a user must affirmatively seek sexually explicit material to view it, on occasion a search not intended to retrieve sexually explicit material may retrieve a link to a sexually explicit site. For example, Schmidt's searches of “Sleeping Beauty,” “Babe,” and “Little Women” produced a handful of links to sexually explicit sites. (Defendant's Exs. 15, 18, 27, 31, 38) This demonstration revealed the inevitable imprecision of search engines—a broad search will almost always return some irrelevant results. In the vast majority of cases, the character of a sexually explicit site will be clear from the entry or link that a search engine returns. Nevertheless, there is potential for occasional accidental viewing of sexually explicit material. For example, if a user were to view entries in a WebCrawler search using that program's standard format as preset by the manufacturer, he would see no summary of the sites' contents. (Defendant's Ex. 18; Shirky Test., Tr. at 237–38) One of Schmidt's searches of “Sleeping Beauty” returned an entry offering a link to a site containing sexually explicit material; the entry (when viewed apart from other entries on the same page with similar addresses) gave little indication of the site's contents. (Defendant's Ex. 15; Shirky Test., Tr. at 238) It is difficult to know how often accidental viewing can occur, but there is no basis in the record for concluding that a user not seeking out sexually explicit material on the Internet will encounter it with any particular frequency.

2. The Availability of Sexually Explicit Content

Although Schmidt's demonstration focused on the World Wide Web, sexually explicit content is available on the Internet through almost any form of Internet communication. Yet there is no evidence that sexually explicit content constitutes a substantial—or even significant—portion of available Internet content. While it is difficult to ascertain with any certainty how many sexually explicit sites are accessible through the Internet, the president of a manufacturer of software designed to block access to sites containing sexually explicit material testified in the Philadelphia litigation that there are approximately 5,000 to 8,000 such sites, with the higher estimate reflecting the inclusion of multiple pages (each with a unique URL) attached to a single site. (Stipulated Portions of Record in ACLU/ALA (“Stipulated Record”), Ex. M, at 139–40 (Testimony of Ann W. Duvall in ACLU/ALA )) The record also suggests that there are at least thirty-seven million unique URLs. (Galligher Test., Tr. at 144) Accordingly, even if there were twice as many unique pages on the Internet containing sexually explicit materials as this undisputed testimony suggests, the percentage of Internet addresses providing sexually explicit content would be well less than one tenth of one percent of such addresses.

It is not disputed that some of the sexually explicit materials that the CDA attempts to keep away from minors originates abroad. This is not surprising inasmuch as forty percent of all host computers are located outside the United States. (Joint Stip. ¶ 3) Although only a tentative approximation is possible, the record suggests that as much as thirty percent of the sexually explicit material currently available on the Internet originates in foreign countries. (Stipulated Record, Ex. L, ¶ 41; id. Ex. M, at 161–62 (Duvall Test.))

D. The Development of Blocking Tools and Labeling Schemes

As the Internet has become accessible to more households, several commercial on-line services and software companies have developed features and packages designed to enable parents to limit children's exposure to potentially inappropriate Internet material. *932 For example, America Online, Prodigy, and Microsoft Network, which permit their subscribers to obtain access to Internet material, offer parental control options free of charge to their members. (Joint Stip. ¶ 67) America Online, for example, allows parents to establish a separate account for their children limited to the service's own proprietary content. (Id.) In addition, at least one type of screening software, SurfWatch, has a feature allowing parents to block access to all Internet sites except for those that parents choose to make available to their children. (Stipulated Record, Ex. M, at 131 (Duvall Test.))

The Government offered testimony and a demonstration regarding SurfWatch (configured to act as a screening tool, rather than to block all Internet access) and a second type of screening software, Cyber Patrol. SurfWatch and Cyber Patrol maintain lists of sites known to contain sexually explicit material; when operating while a user attempts to retrieve Internet material, access to sites identified on their programs will be blocked. In addition, the programs block access to sites whose URLs contain particular character patterns or words, such as “xxx” or “sex,” and block any searches including those character patterns or words.

Because of the constant change in the number and location of Internet sites, both SurfWatch and Cyber Patrol offer regular subscription or update services. But even where a parent has properly installed screening software and the software is operational (and configured to block access to certain sites rather than to the entire Internet), it is possible to retrieve some sexually explicit material. The Government's witness was able to run searches using “Babe” and “Little Women” as key words with screening software running in the background. As with searches performed in the absence of screening software, the searches returned links to sexually explicit materials. Some of the links were not blocked by the screening tool. In addition, the Government's witness obtained access to sexually explicit material by directly entering URLs obtained from earlier searches conducted without blocking software in the background. The record also shows that blocking software is not widely owned by or used in households with access to the Internet: nearly seventy percent of SurfWatch's 1,500 subscribers are schools rather than individual households. (Id. at 163–65) Other efforts to assist parents in filtering and screening material that their children can view on the Internet are under way. The World Wide Web Consortium (“W3C”) has launched the Platform for Internet Content Selection (“PICS”) to develop technical standards for attaching electronic ratings to Internet addresses. (Joint Stip. ¶¶ 47–49; Stipulated Record, Ex. J., at 1; id. Ex. G, at 2–3 (Declaration of Albert Vezza in ACLU/ALA )) When the system is fully implemented, PICS-compatible client software (including browsers, newsgroup readers, and mail readers); Internet service providers; and commercial on-line services will be able to detect PICS tags and block content based on how a parent has configured the software. (Joint Stip. ¶ 48; Stipulated Record, Ex. G., at 3 (Vezza Decl.)) PICS will thus enable parents to design from an array of categories blocking criteria that suit the parents' values or needs. The PICS program envisages both rating by content providers and rating by third parties. (Joint Stip. ¶ 48) The vast majority of Internet sites currently remain unrated. Nevertheless, Microsystems Software, Inc. (which manufactures Cyber Patrol) introduced a PICS ratings server in February 1996. (Id. ¶ 54) Cyber Patrol is itself now PICS-compatible; it can screen out material based on its PICS tag. (Id.) In addition, Microsoft released the first PICS-compatible Web browser, Internet Explorer 3.0, on May 28, 1996. The browser allows parents to block children's access to all unrated Internet sites and to specify appropriate levels of violence or nudity at rated sites. (Hoffman Supp.Decl., Ex. C, at 1–3) In addition to PICS tags, the Government's expert witness, Dr. Dan Olsen, testified that content providers wishing to transmit or make available material potentially falling within the scope of the CDA could develop a general practice of inserting a “tag” or “label”—a string of characters, such as “–L18” (for “not less than 18 years”)—into the address or name of a particular site so as to clearly identify the site as unsuitable for minors. To transmit or gain access to Internet content, a user must specify a textual name: one cannot send e-mail without an e-mail address or the name of a mailing list; post an article to a newsgroup without specifying the name of the group; participate in the Internet Relay Chat without specifying a “channel”; or access a file without its address. (Olsen Decl. ¶¶ 22–26) Accordingly, content providers using all significant modes of Internet communication could use a tag to identify their content as “covered” content. For example, when a sender transmits an e-mail message, the message is accompanied by the sender's address, which contains a “user name” identifying a particular user and a “domain name” assigned to a computer or set of computers.[19] (Olsen Decl. ¶¶ 25, 60) If the string –L18 were added to the domain name, all e-mail originating from that site—regardless of the particular user who transmitted it—would be identified as containing material falling within the scope of the CDA.[20] In the alternative, a particular user name—rather than a domain name—could contain the “–L18” tag; only e-mail originating under that user name would be tagged.[21] Finally, a tag could be placed in a textual subject line, so as to identify only particular messages (rather than all e-mail sent under a certain user name or from a certain computer) as containing content potentially within the scope of the CDA. (Id. ¶¶ 60–62)

Similarly, a tag such as “–L18” could be added to the name of a newsgroup; an individual user wishing to post an article potentially falling within the scope of the CDA to a newsgroup that does not as a general matter contain such material could insert a tag in the subject line accompanying the article. (Id. ¶¶ 64–65) A tag could also be placed in the name of an IRC channel.

Turning to means of making files available for retrieval or viewing by remote users—using an FTP, gopher, or Web server—content providers could insert a specific tag such as “–L18” in a domain name or site name. Thus, as the Government's expert witness testified, an owner of a Web site named “www.cyberporn.com” could rename the site “www–L18.cyberporn.com”. (Id. ¶ 51) If a site only contained specific files falling within the scope of the CDA, a content provider could identify those files by adding a tag to the name of the directory in which the file resides or to the file name itself. That is, a file identified with the address “http://www.adult.com/picturel.html/” could be renamed “http://www.adult.com/picture1–L18.html/”; in the alternative, a content provider could place all covered files within a specific directory, such as “http://www.adult.com/pictures–L18/”. (Id. ¶¶ 51–54) A content provider who did not wish to tag an entire file available on a Web server as unsuitable for minors could place a tag within the HTML source code of the file, thus identifying a particular section as subject to the CDA. (Id. ¶ 58) In any of these approaches, tagging content is, in a technical sense, a trivial act. (Id. ¶¶ 59, 62; Stipulated Record, Ex. B, at 56 (Testimony of Scott O. Bradner in ACLU/ALA ))

There is an alternative means to shield minors from sexually explicit content available uniquely to content providers on the World Wide Web: verification of a user's “adulthood” before allowing him access to a site. A content provider operating a Web server can create and display an electronic form to retrieve information from a user visiting the Web site; after processing the information by using a program such as a Common Gateway Interface (“cgi”) script, the server could grant or deny access to the site. (Shirky Decl. ¶ 21) Not all content providers who make material available on the Web, however, can use programs such as cgi scripts; for example, commercial on-line services such as America Online and CompuServe provide subscribers with the opportunity to post content by configuring their own Web pages but do not permit subscribers to use cgi scripts. (Olsen Test., Tr. at 345) For Web content providers who lack access to cgi scripts, there is no means of age verification.

Although some Web providers can query the user of a site for a credit card number, the cost of verification is significant, ranging from sixty cents per transaction to more than a dollar per transaction. (Id. at 341–42) To take advantage of adult access code or adult identification code verification, a content provider would either have to establish and maintain a registration and verification system (or hire someone else to do so) and issue access codes to users—after verifying their ages—or associate with one of several adult verification services, such as Adult Check, Adult Verification System, First Virtual, Validate, or VeriSign. (Olsen Decl. ¶ 86 & Ex. I; Schmidt Test., Tr. at 203–14; Defendant's Exs. 6, 7, 8 & 9) Although neither of the Government's expert witnesses had any first-hand familiarity with adult verification services, advertising materials suggest that an adult can obtain an identification number from a particular service and access any site registered with the service. For example, a user can register with Adult Check for an annual fee of $9.95; when the user attempts to access any site registered with Adult Check, the user is prompted to enter an Adult Check identification number that is checked against the service's database. (Defendant's Ex. 6, at 1) If the number is valid, the user is automatically admitted to the site. (Id.) Although most verification services do not charge content providers to register their sites (Id. Exs. 6–8), at least one service does impose a fee on site owners registered with it. (Id. Ex. 9, at 1)

Having explored various means of Internet communication, the availability and accessibility of sexually explicit content, the development of blocking software and rating schemes designed to enable parents to shield their children from inappropriate material, and the potential for tagging and verification procedures that content providers can themselves employ in an effort to shield minors from sexually explicit content that they provide, we turn to the governmental regulation in question.

Footnotes

  1. It became clear from the testimony that moderated newsgroups are the exception and unmoderated newsgroups are the rule.
  2. The evidence adduced at the hearings provided detail to this paragraph of the parties' stipulation. See Findings 95 to 107.
  3. Testimony adduced at the hearing suggests that market forces exist to limit the availability of material on-line that parents consider inappropriate for their children. Although the parties sharply dispute the efficacy of so-called "parental empowerment" software, there is a sufficiently wide zone of agreement on what is available to restrict access to unwanted sites that the parties were able to enter into twenty-one paragraphs of stipulated facts on the subject, which form the basis of paragraphs 49 through 69 of our Findings of fact. Because of the rapidity of developments in this field, some of the technological facts we have found may become partially obsolete by the time of publication of these Findings.
  4. This membership is constantly growing, according to the testimony of Albert Vezza, Chairman of the World Wide Web Consortium.
  5. Mr. Bradner is a member of the Internet Engineering Task Force, the group primarily responsible for Internet technical standards, as well as other Internet-related associations responsible for, among other things, the prevailing Internet Protocols. He is also associated with Harvard University.
  6. Dr. Olsen chairs the Computer Science Department at Brigham Young University in Provo, Utah, and is the recently-appointed Director of the Human Computer Interaction Institute at Carnegie-Mellon University in Pittsburgh, Pennsylvania.
  7. The term "information content provider" is defined in § 509 of the CDA, at the new 47 U.S.C. § 230(e) (3), as "any person or entity that is responsible, in whole or in part, for the creation or development of information provided through the Internet or any other interactive computer service."
  8. By "verification", we mean the method by which a user types in his or her credit card number, and the Web site ensures that the credit card is valid before it allows the user to enter the site.
  9. InterNIC is a naming organization, not a regulator of content. InterNIC and two other European organizations maintain a master list of domain names to ensure that no duplication occurs. Creators of Web sites must register their domain name with InterNIC, and the agency will instruct the creator to choose another name if the new Web site has the name of an already-existing site. InterNIC has no control over content on a site after registration.
  10. This paragraph and the preceding paragraph also illustrate that a content provider might store its own material or someone else's on a caching server. The goal saving money and time is the same in both cases.
  11. We use the term “content” to refer to any text, data, sound, program, or visual image transmitted over or made available for retrieval on an interactive computer service.
  12. While § 223(d) regulates more than the content of Internet communications, we focus mainly on the range of tools and services available to individuals with Internet access, recognizing that the vast majority of content available through the use of an interactive computer service is in fact available on the Internet.
  13. A “modem” (a contraction of “modulator” and “demodulator”) is a device that translates digital information into a signal for transmission over a telephone line (“modulation”) and translates a signal received over a telephone line into digital information (“demodulation”).
  14. We use the term “content provider” to refer to any Internet “speaker”—that is, a user who transmits or makes available any content over the Internet. Although the term is not used in the statutory provision at issue, “information content provider” is elsewhere defined in the CDA as “any person or entity that is responsible, in whole or in part, for the creation or development of information provided through the Internet or any other interactive computer service.” Pub.L. No. 104–104, § 509(e)(3), 110 Stat. at 139 (to be codified at 47 U.S.C. § 230(e)(3)). The phrase serves as a reasonable shorthand for the category of individuals targeted by the CDA—persons who send or display Internet content. See S.Conf.Rep. No. 230, 104th Cong., 2d Sess. 188 (1996) (stating that § 223(d)(1) applies to “content providers”).
  15. We use the term “software” to refer to the combination of programs and procedures that serve as instructions to the computer. The term is often used in contrast with “hardware,” which refers to a computer system's physical elements.
  16. To locate files available for copying, a user can contact an “Archie” server—a remote computer capable of searching directories for file names containing a particular string of characters on FTP servers permitting anonymous retrieval. (Hoffman Decl., Ex. 4, at 180–90) (Hoffman Decl., Ex. 3, at 1–2, 5; id. Ex. 4, at 187; Joint Stip. ¶ 29)
  17. As with FTP servers, there are tools available for locating menus or items containing a certain string of characters: a “Veronica” server is capable of searching menus on all gopher servers, while “Jughead” is an aptly named tool for searching menus on only a single server. (Galligher Test., Tr. at 124; Hoffman Decl., Ex. 3, at 5; id. Ex. 4, at 191–92)
  18. Most of these services do not charge users for search requests and are sustained primarily by advertising revenues. (Galligher Test., Tr. at 136–37)
  19. In the example [email protected], “smith.com” would constitute a domain name.
  20. Following the example above, all e-mail would originate from the domain smith-L18.com.
  21. In the example above, material would originate from the address [email protected].