[NOTE: This analysis applies no longer to S.314, which was heavily amended, then folded into the Senate telecom reform bill, S652, on Mar. 23, 1995. It does however apply to the House version of the bill, H.R.1004, which has not been amended and preserves the original text. - mech@eff.org 3/28/95] ELECTRONIC FRONTIER FOUNDATION POLICY PAPER: THE EXON BILL, S. 314, 1995 ------------------------------------------------------------------------ This is the first in a series of documents by EFF Policy Fellows. While they do not necessarily reflect the position of the EFF Board of Directors, EFF feels these Policy Papers present interesting viewpoints, worthy of consideration and study. Electronic Frontier Foundation Policy Paper: Exon Amendment (S. 314) February 28, 1995 Contact: David G. Post (dpost@eff.org) EFF Policy Fellow Summary of Proposed Changes The changes proposed in Section 2 of the Exon Amendment (the "Amendment") fall into 2 categories. 1. A number of the proposed changes would expand existing criminal liability by making certain acts that are currently illegal if done using the telephone illegal if done using a "telecommunications device." A. Current law makes it illegal to "make" an "obscene, lewd, lascivious, filthy, or indecent" comment by means of a telephone. The Amendment would extend criminal liability to anyone performing such an act by means of a "telecommunications device." Sec. 2(a)(1)(A). B. Current law makes it illegal to "make" a telephone call "without disclosing [your] identity" and "with intent to annoy, abuse, threaten, or harass any person at the called number." The Amendment would extend liability to anyone who performs a similar act "utiliz[ing] a telecommunications device." Sec. 2(a)(1)(C). C. Current law makes it illegal to "make repeated telephone calls" to "harass any person at the called number." The Amendment would extend criminal liability to anyone who "repeatedly initiates communication with a telecommunications device" for that purpose. Sec. 2(a)(1)(D). D. Current law makes it illegal to "knowingly permit a telephone facility under [your] control" to be used for any of the above illegal purposes. The Amendment would extend liability to anyone who knowingly permits a "telecommunications facility" to be used for such purposes. Sec. 2(a)(2). E. Current law makes it illegal to "make an indecent communication for commercial purposes . . . available to any person under 18 years of age" by means of a telephone. The Amendment would extend criminal liability to those making such communications available "by means of a telecommunications device." Sec. 2(a)(4). 2. A number of the proposed changes would expand existing criminal liability by making a new class of persons liable for the illegal acts of others. A. Current law makes it illegal to "make" an "obscene, lewd, lascivious, filthy, or indecent" comment by means of a telephone. The Amendment would extend criminal liability to anyone who "transmits or otherwise makes available" such a comment. Sec. 2(a)(1)(B). B. Current law makes it illegal to "make an indecent communication for commercial purposes . . . available to any person under 18 years of age" by means of a telephone. The Amendment would extend criminal liability to anyone who "knowingly transmits or knowingly makes available" such communications. Sec. 2(a)(4). Objections to the Proposed Changes A. The extension of liability to acts performed via "other telecommunications devices" It may seem logical to propose that a certain act that is now illegal if accomplished via a telephone should be illegal if accomplished via electronic mail or other telecommunications devices. It is a logic that breaks down on closer examination. 1. Existing law embodies the principle that persons should not be permitted to exploit the capability of this medium to impose harmful or harassing communications on unwilling listeners. By the same token, private, consensual communication is, simply, none of the government's business (even if that communication might otherwise be deemed "obscene, lewd, lascivious," etc.). The telephone is both a point-to-point medium and an "intrusive" medium. Calls are directed to individual recipients who are generally unable (notwithstanding a measure of protection available through Caller ID or similar technology), to avoid calls placed directly to their telephone number, even if they are unwilling to participate in the communication. Existing law thus criminalizes only those communications that are non-consensual -- because only the sender and the recipient are participants in the call (and no one will be in a position to complain if the two participants consensually engage in the communication). The ban in existing law on obscene telephone calls, or repeated anonymous and "harassing" telephone calls, protects the unwilling recipient from being subjected to communication that he/she finds offensive and cannot avoid. 2. Any extension to the world of electronic communication should sweep no more broadly than to cover personally targeted, non-consensual, point-to-point electronic communication, because only in that context does the concept of the "unwilling listener" make sense. "Offensive" communications that appear in electronic discussion groups, chat "rooms," Usenet groups, listservers, and the like are not akin to personally directed telephone calls intruding upon an unwilling listener. These contexts are more closely analogous to voluntarily-entered communities, where the unwilling listener can always exit -- avoiding exposure to the unwanted communication -- if the communications are offensive or otherwise annoying. B The extension of liability to persons who "transmit or otherwise make available" specific kinds of communications 1. The extension of liability to persons who "transmit or otherwise make available" specific kinds of communications will have a substantial and detrimental effect on robust communication on individual networks and the Internet. The potential breadth of this aspect of the Amendment is breath-taking, given the decentralized architecture of the Internet. Any given message travelling over the Internet may be "transmitted or otherwise made available" by dozens of intermediate networks; the Amendment, on its face, would appear to make these intermediate network administrators criminally liable for the content of messages "transmitted" by their systems. Additionally, any of the possibly thousands of network administrators who might make an individual Usenet group available to network users would, it appears, be liable should e.g., any "indecent" communication appear in the communications within that group. The risks for network administrators under this regime would be substantial even if it were possible to pre-screen all communications prior to "making them available," and even if there were clear definitional boundaries such that one could easily determine ex ante what was, or was not, an "obscene, lewd, lascivious, filthy, or indecent" communication. Given the virtual impossibility of the former and the absence of the latter, the risks are further exacerbated. It is inevitable that network administrators and others in control of telecommunications facilities will take steps to minimize these enormous risks, steps certain to include a refusal to transmit communications from discussion groups or other Internet sites that may contain material that may be deemed to fall into one or more of the prohibited categories. This will necessarily chill the robust discussion that has characterized the online environment up to this point by making unavailable an enormous range of communication outside of the prohibited categories -- communication squarely within the protection of the First Amendment. 2. Imposition of liability only on those who "knowingly" transmit or make available prohibited communications will not cure this deficiency. The term "knowingly" does not consistently refer to a single, well-defined scienter (mental state) standard. It is unclear, and will continue to be unclear until courts give a definitive construction to this term in this specific context, whether, for example, (a) knowledge that a particular newsgroup is being made available to network users, or (b) knowledge that the newsgroup is being made available and is unmoderated, or (c) knowledge that the newsgroup has carried prohibited communications (frequently?) in the past, or (d) knowledge that a specific communication was carried in the newsgroup, or (e) knowledge that the content of the specific communication fell within the category of prohibited communications, will be sufficient to impose liability on a network administrator in the event that an "obscene" communication appears among the many non-prohibited communications in the group. Perhaps more significantly, a scienter standard for intermediaries will, perversely, discourage precisely the kind of monitoring of communication that offers users protection against being subjected to unwanted communications. Network administrators will have the incentive to avoid all monitoring or pre-screening of messages in order to avoid being deemed to have "knowledge" of a specific prohibited communication. Providers of transmission services should instead be given maximum flexibility to set local standards for the classes of messages they are willing to carry, to perform the monitoring they deem appropriate to enforce those standards, and incentives should be provided to encourage them to communicate the nature and scope of those actions so as to enable users to make informed decisions about the kind of communication to which they wish to be exposed. A Counter-Proposal For all of its obvious and serious flaws, the Exon Amendment provides those interested in the free flow of information in electronic communities with the opportunity to re-think the question of content regulation in those communities and to design standards that will provide protection against harm to the unwilling listener without sacrificing the vigorous debate that has come to characterize those communities. We offer the following principles that, if embodied in legislation, would better balance these competing needs. 1. The Internet -- and, increasingly, individual constituent networks -- are global in scope, presenting difficult jurisdictional questions concerning the proper law to apply in particular instances. Although federal legislation cannot alone solve these conflict-of- laws questions, a properly drafted federal statute that expressly pre-empts conflicting state legislation could be an important first step towards relieving network administrators and others responsible for the Internet infrastructure of the burden of complying with overlapping and mutually inconsistent State law. 2. The principle of individual responsibility for harmful acts initiated by the individual should be a centerpiece of any legislative initiative in this area. Legal remedies for any harm resulting from electronic communications -- for copyright infringement, harassment, defamation, or the like -- should focus upon the person or persons responsible for initiating the communication. In order to preserve the free flow of information within and across electronic communities, persons who do no more than transmit or forward messages should be immune from liability. Only if such persons take affirmative steps to solicit, sponsor, promote, or adopt the communication in question as their own, or if, by their actions, they make it more difficult for the recipient to identify the originator of the offending communication, should liability attach to their actions. The following language, broadly adopted from the draft of an Electronic Communications Forwarding Act proposed by EFF (and others) several years ago, would be a useful starting point for legislative action: Any person who, using an instrumentality of interstate commerce, forwards or transfers such electronic communication between the originating party and authorized recipients, not be subject to criminal or civil penalty nor liable to any other person for damages or equitable relief of any kind, provided that such person does not solicit, sponsor, promote, or adopt such communication as its own, and provided further that such person does not alter the identifiability of the sender of such communication. 3. The free flow of information in electronic communities can and should accommodate adequate protection for the unwilling listener against intrusive, personally targeted, non-consensual, point-to-point electronic communication, and we support the development of uniform standards in this area. Such standards should be based upon the principle that message recipients can avoid exposure to any such offending messages, but only if the messages are properly labelled and their originator(s) properly identified; hence liability should attach only where the sender takes affirmative steps to shield the recipient from information about the identity of the sender, or to mislabel or misrepresent message content, in such a way as to make it unreasonably difficult for the recipient to avoid communication that he or she deems offensive. 4. The means by which definitional questions regarding the boundaries of communications deemed "obscene" or otherwise unlawful must respect the right of communities to self- determination and self-governance to the maximum extent possible. Outside of the electronic context, reference to prevailing "community standards" in making these determinations attempts to balance the right of residents in specific communities to exercise control over the material that enters their communities, against the right of creators and distributors of diverse material to operate freely in those communities where that material is not offensive to prevailing standards, so as to encourage diversity of expression and to prevent imposition of a uniform, "lowest common denominator" standard across different communities. The world of electronic communication uniquely offers individuals the ability to associate into virtual communities with others who share their sensibilities regarding material that other communities might deem indecent, obscene, lascivious, or lewd. The "boundaries" between these online communities are passwords and warning screens, that enable the nature of the communications permitted (and prohibited) in the community in question to be clearly communicated to all those who wish to join the community in question. Accordingly, we believe that the best source of a definition regarding what constitutes "obscenity", for purposes of determining when the law should intervene to prohibit electronic distribution of materials, is the standard set by the users who, collectively, set the rules applicable to any particular online community. Where the nature of the materials is clearly disclosed on warning screens encountered as the users access the BBS system, those who sign on -- who voluntarily join the community - - have already determined that the materials in question do not violate their own sensibilities. If the operators of a BBS system were to post materials that violated the collective standards of that user community, the community in question could quickly correct things by voting with their modems to go elsewhere. 5. One problem, ostensibly addressed by the Amendment, remains: how to insure that adult-oriented material does not find its way into the hands of minors? This problem can arise in two broad contexts in the online environment: membership-restricted forums (e.g., adults-only BBSs) and open, unrestricted forums (e.g., WWW sites). Regarding the former, EFF supports imposition of a duty on system administrators to take reasonable steps to collect information concerning the age of persons attempting to access the material, analogous to the duties of operators of "900" telephone services to obtain similar information regarding callers. Regarding the latter, we believe that parental control over minors' access can best be achieved through a combination of standardized header codes and "warning labels" in conjunction with software-based filtering devices, and we will continue to support efforts by the online community itself to develop and implement such standards.