INTRODUCTION: IS A FEDERAL STANDARD OF LIABILITY FOR INTERNET SERVICE PROVIDERS NECESSARY?
I.   STATE OF THE CURRENT REGIME: THE ENACTMENT OF THE PUBLISHER/DISTRIBUTOR DISTINCTION
II.  BLUMENTHAL v. DRUDGE: IS ISP IMMUNITY A WORKABLE STANDARD FOR INTERNET PUBLISHING
III. "DISTRIBUTOR" LIABILITY FOR INTERNET SERVICE PROVIDERS: HOW IT CAN WORK AND WHEN IT SHOULD APPLY
IV. CAN ISPs FUNCTION EFFECTIVELY IF HELD TO "PUBLISHER LIABILITY"?: INSTANCES WHEN IT SHOULD AND THE PROTECTION OF THE ACTUAL MALICE STANDARD
V.  INTERNET JURISDICTIONAL PROBLEMS: WHY A FEDERAL STANDARD IS NECESSARY
CONCLUSION

Few things in life are more actively sought, or treasured, in today's society than immunity from prosecution or liability. Whether one is a corporate director, a foreign diplomat, or a testifying potential felon-the rationale for granting the privilege is usually the same: the freedom from liability encourages a person to act in a manner in which she would not without the privilege. The institution grants the privilege because of its value judgment that the behavior the privilege is thought to encourage is more valuable than the potential costs of granting the privilege. These judgments are made everyday by legislators, courts, agencies, and other powerful bodies. Congress made such a decision in 1996 by enacting the Communications Decency Act of 1996 ("CDA"),[1] which strongly limited the potential liability of internet service providers ("ISPs") for information content created by third parties that can be accessed via the Internet from their servers.[2]

In a series of decisions since the enactment of the CDA, the federal courts have interpreted the Act to bar virtually all liability of ISPs for providing access to defamatory information created by third parties.[3] The vast majority of ISPs are in the business of selling access to third parties and publish virtually no information content of their own, and those third parties are likely to have meager funds of their own to satisfy a defamation judgment; therefore, the effect of the CDA has been to transform the Internet into an almost liability-free zone for libelous content. As traditional print media moves more and more toward the Internet, such a regime can eventually do considerable harm not only to those whose reputations and livelihood are endangered by libelous statements, but also to the potential of the Internet as a reliable, easily accessible, and inexpensive means of communication for the reporters of tomorrow as well as for existing print and broadcast media. The development of a more inexpensive and accessible print media has been one of the Internet's greatest innovations, and one for which the proliferation of ISPs is largely responsible.

Though the rapid development of the Internet has created a fertile ground for legal innovation, more often than not legislators and courts have sought to address this relatively new medium by attempting to squeeze it into precedents and paradigms better suited to older forms of communication, technology, and media.[4] Part I of this article looks back at the courts' initial efforts at addressing defamation via the Internet. From the start the courts attempted to fit the role of the ISP into the common law's categorizing of print media as either "publishers" or "distributors" of information. One court's misstep in overextending the liability of one ISP to that of a publisher led Congress to sweep ISP liability into the CDA in an attempt to give ISPs more tools to regulate content they consider offensive.[5] Part II then takes a look at the most recent major Internet defamation case, Blumenthal v. Drudge, and how it cemented the current regime, which interprets Congress's prohibition of ISP liability for "publishers" as a prohibition against virtually all liability and not a narrowing of ISP liability to that of a "distributor."[6]

The potential ramifications for allowing ISPs to be liable as "distributors" of third party defamatory content are addressed in Part III. In particular, the section focuses on whether a notice-based liability regime could function on the Internet without either compromising the medium's main advantages or chilling free speech. Part IV takes into account these same concerns while arguing for a narrow application of "publisher" liability to those few ISPs that fulfill a content-providing role akin to that of a publisher of a newspaper or magazine. By examining the Drudge case directly, this comment demonstrates how traditional safeguards in defamation law like the "actual malice" standard for public figures will severely limit the chances of these ISPs to fall victim to unwarranted liability, and that a federal exemption from "publisher" liability for ISPs was never really necessary.

Defamation law is one field where Congress could effectively legislate a federal standard based on prior precedents with minor tweaking without taking the sledgehammer tactic of providing blanket immunity to those who provide access to the defamers of tomorrow. A federal standard is necessary due to the unique jurisdictional concerns of the Internet that are examined in Part V. Congress should reform the CDA to create a more even playing field between Internet media providers and broadcast and print media providers so that the former (just as much as the latter) may be found liable as "distributors" when they fail to remove defamatory content of which they are aware. In addition, if an ISP commissions and edits information content in a manner analogous to newspaper or book publishers, it should be held to the stricter "publisher" standard of fault as well.

The portion of the "protection for 'good samaritan' blocking and screening of offensive material"[7] provision of the CDA at issue in the Internet defamation cases states simply that "[n]o provider or user of an interactive computer service shall be treated as the publisher or speaker of any information provided by another information content provider."[8] Congress enacted this amendment specifically to counteract the decision in Stratton Oakmont, Inc. v. Prodigy Services Co.[9] that held an ISP liable as a "publisher" for defamatory information content created by a third party.[10] Congress feared this decision would discourage ISPs from taking active steps to monitor and remove offensive content from public view.[11] In Stratton Oakmont the term "publisher" was used to differentiate entities like newspapers and magazines that exercise editorial control and judgment from "distributors" of information like bookstores that are only held liable if "they knew or had reason to know of the defamatory statement at issue."[12] Considering that the provision was enacted to overturn that case, one would think that the courts would find the meaning of the word "publisher" within the context of that distinction. The Fourth Circuit, however, in Zeran v. America Online, Inc., chose to interpret it as applying in a general sense to all "publication" of information.[13] Such an interpretation, in the words of the Zeran court, creates a "federal immunity to any cause of action that would make service providers liable for information originating with a third party user of the service."[14] This immunity only flows from the CDA, however, if one interprets "publisher" in this general sense and not in the context of the publisher/distributor fault distinction used in the very case that Congress sought to address.

Stratton Oakmont involved a claim that the ISP Prodigy Services was liable as a publisher for an allegedly defamatory statement made by an anonymous user on Prodigy's "Money Talk" computer bulletin board about the plaintiff-a securities investment banking firm named Stratton Oakmont, Inc.[15] The Stratton Oakmont court relied on caselaw that holds "distributors" and "publishers" of defamatory matter to different standards of fault.[16] While a publisher (for example, a newspaper, TV network, or magazine) is subject to liability as if it had published the third party defamatory material itself, a distributor (for example, a bookstore or a newsstand) can be liable only if it knew or had reason to know of the defamatory material.[17] The rationale for treating distributors differently is that they are more like "passive conduit[s] and will not be found liable in the absence of fault."[18]

In an earlier case involving another large ISP (CompuServe) and a similar defamation claim regarding a bulletin board posting, Cubby Inc. v. CompuServe Inc.,[19] the court held that, due to the ISPs lack of editorial control over the publication of articles on the bulletin board, it was the functional equivalent of a news vendor and thus could only be held liable under a distributor's standard.[20] The Stratton Oakmont court distinguished Prodigy's conduct from CompuServe's by finding that the former maintained enough editorial control over the articles posted on the "Money Talks" bulletin board to merit being held to a publisher's standard of fault.[21] Prodigy had promised the public that it controlled the content on its computer bulletin boards by deleting language it deemed to be offensive, and had done so through the use of automatic software and by promulgating guidelines that directed the Board managers (called "Board Leaders") to remove offensive language and content.[22] Through its own conduct, policies, and technology, Prodigy had thus "altered the scenario and mandated the finding that it is a publisher."[23]

In the wake of Stratton Oakmont it seemed clear that in defamation actions involving online speech the courts would hold service providers who took a largely "hands-off" approach to be generally free from liability as "distributors," while ISPs that monitored and edited content would be held to the stricter "publisher" standard. The court's holding seemed to be at least partly motivated by its own preference for the former approach on policy grounds. For example, it stated that "Prodigy's current system . . . may have a chilling effect on freedom of communication in Cyberspace, and it appears that this chilling effect is exactly what Prodigy wants, but for the legal liability that attaches to such censorship."[24] The court's concern about this "chilling effect" is in keeping with the policy rationale behind modern defamation law that attempts to balance society's interest in providing "breathing space" for robust public debate against the interest of the individual in protecting himself from unfair reputational harm.[25]
The CDA is not suited to address those competing interests because it was not designed with them in mind. In the "Policy" section of the Act, Congress states specifically that it intends to "remove disincentives for the development and utilization of blocking and filtering technologies that empower parents to restrict their children's access to objectionable or inappropriate online material . . ."[26] Thus, the main concern of Congress in enacting the Communications Decency Act was not to absolve ISPs of liability, but to prevent the Internet from becoming a "red light district" and to "extend the standards of decency which have protected users to new telecommunications districts."[27] The crux of the Act was the provisions prohibiting "the knowing transmission of obscene or indecent messages to any recipient under 18 years of age."[28] Not only would those who actively transmitted or knowingly made available "indecent" materials to minors have been civilly or criminally liable under the Act, but so would Internet information content providers (in other words, ISPs) who "knowingly [permitted] any telecommunications facility under [their] control to be used for an activity prohibited [by the Act] with the intent that it be used for such activity . . ."[29] Because one could knowingly permit such actions by doing nothing to monitor them, Congress was clearly trying to encourage information content providers to self-regulate the activities and speech of people who stored information on their servers. To that end Congress enacted the Good Samaritan provisions in the CDA.[30]

In discussing the purpose beyond the "publisher" clause of § 230(c) Congress said nothing about the potential liability of information content providers for libel. What it did state was that the "Good Samaritan" protections from civil liability were intended to aid those ISPs who took "actions to restrict or to enable restriction of access to objectionable online material" and to "overrule Stratton-Oakmont v. Prodigy and any other similar decisions [that] have treated such [internet service] providers and users as publishers or as speakers of content that is not their own because they have restricted access to objectionable material."[31] Thus, one can see Congress's clear intent was to protect and encourage the actions of ISPs that take active steps to monitor and remove objectionable content, not to protect them from liability when they knowingly choose not to remove objectionable content. Stratton Oakmont uses the term "publisher" in no other context than to distinguish those who have control over editorial content from "distributors" who have no such control; and Congress specifically sought to overturn that decision. Though Congress did not state so explicitly in the legislative history, it seems implicit that Congress meant to preserve the distinction and retain the less strict "distributor" liability for known content. To find otherwise would appear at cross-purposes with Congress's intention to encourage active ISP monitoring and enhance user control in order to "empower parents to restrict their children's access to objectionable or inappropriate online material;" and to "ensure vigorous enforcement of Federal criminal laws to deter and punish trafficking in obscenity, stalking, and harassment by means of computer."[32]

Even though Congress specifically pointed to its desire to overrule Prodigy in the legislative history, the first major case to apply the provision in a defamation action inferred a definition of the term "publisher" that was broader than the one traditionally employed in defamation law to distinguish "publishers" from "distributors."[33] In Zeran v. America Online, the plaintiff Zeran brought a negligence action against America Online alleging that the ISP had unreasonably delayed the removal of defamatory statements posted by an unidentified third party, refused to post any retractions, and failed to screen its content for subsequent postings regarding Zeran.[34] Zeran clearly had suffered harm to his reputation-the anonymous poster had sent out repeated phony messages advertising T-shirts ridiculing the Oklahoma city bombing and directing those "interested" to contact a phone number that was in fact Zeran's home number.[35] Since Zeran ran his business out of his home, the predictable angry calls and death threats that resulted from the messages caused damage to his livelihood as well.[36] The actions of the poster clearly would seem to fit the requirements of the tort of defamation: publication of defamatory communications that injure the plaintiff's reputation by diminishing the esteem in which the plaintiff is held, and exciting derogatory feelings against him.[37] It is generally sufficient that the communication "would tend to prejudice the plaintiff in the eyes of a substantial and respectable minority."[38]

In light of the CDA provisions eliminating the potential for ISPs to be found liable as "publishers" of third party conduct, Zeran sought to hold AOL liable for negligence as a distributor of the material.[39] The court circumvented the publisher/distributor dichotomy, however, by stating that when Congress was speaking of "publishers," it meant to include both publishers and distributors.[40] Thus, it held that Congress was referring to the use of the term "publication" in defamation law as it is used generally to describe the "communication intentionally or by a negligent act to one other than the person defamed."[41] Though the court correctly noted that Stratton Oakmont and Cubby "do not . . . suggest that distributors are not also a type of publisher for purposes of defamation law,"[42] this interpretation fails to take into account Congress's specific overruling of Stratton-a case that only addressed media which belong to the "publisher" subset, and not to all published material.[43]

Since Zeran dealt with a largely unregulated bulletin board, there was no need for the court to take into consideration the dual nature of America Online as both an ISP and a content provider. America Online (like Prodigy and CompuServe), however, is not a conventional ISP that merely provides the consumer access to the Internet. It also publishes its own and third party content in an area that is only open to its own subscribers.[44] This "user-only" content area has been a key factor in AOL's success as it enables it to attract many neophytes and less computer-literate World Wide Web ("WWW") users to its service.

In 1997, AOL entered into a written license agreement with Matt Drudge, the proprietor of an electronic publication on the World Wide Web entitled the Drudge Report.[45] The site was, and still is, available for free to all with access to the WWW and consists of both links to other news sources as well as Drudge's own personal news report.[46] In return for permission to post content from Drudge in its subscriber-only area, AOL agreed to pay Drudge a monthly royalty fee of $3,000.[47] Under the agreement AOL also reserved the rights to remove content from the site that it deemed to "violate AOL's then standard terms of service,"[48] and to "require reasonable changes to . . . content, to the extent such content will, in AOL's good faith judgment, adversely affect operations of the AOL network."[49]

Thus, one can see that AOL was in fact publishing third party content and reserving the right to edit that content. AOL also issued a press release in which it touted the arrival of The Drudge Report and stated that their agreement will make his reports "instantly accessible to [AOL] members who crave instant gossip and news breaks."[50] The court referred to these facts in expressing its frustration that, due to the CDA, it could not find the provider liable as a publisher of content despite the fact that it is "not a passive conduit like the telephone company" and "has the fight[sic] to exercise editorial control over those with whom it contracts and whose words it disseminates."[51] In light of Zeran and the CDA, the Court felt it was bound by the CDA to defer to Congress's supposed intent to provide immunity from tort liability for third party content to ISPs as an incentive for the providers "to self-police the Internet for obscenity and other offensive material, even where the self-policing is unsuccessful or not even attempted."[52]

Though the court interpreted CDA phrases like "offensive material"[53] and "otherwise objectionable"[54] material to cover defamatory material, it is quite likely that Congress did not intend to give such a broad scope to these terms. In the Senate Conference Report for the bill the "good samaritan" provisions are justified as a means to protect ISPs who take steps to provide parents with the ability "to determine the content of communications their children receive through interactive computer services."[55] How many parents are seriously trying to prevent their children from gaining access to defamatory content, and how would they know it was defamatory in the first place? Defamatory content is only offensive to the person who suffers injury from harm to her reputation, and a relatively limited number of family, friends, and other supporters. It is not by its nature something parents would necessarily feel the need to restrict, or content that would be offensive on its face. In addition, because of the vague and overly broad definitions of "indecent" and "patently offensive" material, the Supreme Court overruled most of the CDA in Reno v. ACLU. [56] Thus, to rule that Congress intended the CDA to address defamation liability requires resorting to a definition of offensive material that the Court found to place "an unacceptably heavy burden on protected speech."[57]

The great irony of the granting of enhanced protection to "good samaritan"[58] information content providers is that it has survived the overturning of the restrictive provisions of the CDA that the provisions were designed to counterbalance. In Reno v. ACLU, the Supreme Court overturned the indecency provisions of the Act on First Amendment grounds stating that they were impermissible content-based restrictions on speech.[59] The Court found it crucial that the regulation in question was "more likely to interfere with the free exchange of ideas than to encourage it" and that society's interest in encouraging that freedom outweighed "any theoretical but unproven benefit of censorship."[60] In light of this decision and the subsequent expansion of ISP immunity from liability in Zeran and Drudge, it appears that we have encountered a regime where, in the words of the Drudge court, ISPs may take "advantage of all the benefits conferred by Congress in the Communications Decency Act, and then some, without accepting any of the burdens that Congress intended."[61] The question then becomes simply: is this a good thing? If it is not, we will need to find a compromise position that will satisfy the balancing test at the heart of modern defamation law-how to "reconcile the interest in reputation with the interest in freedom of speech."[62] In doing so we will have to keep in mind how this ancient tort can exist in the rapidly changing worlds within Cyberspace, while recognizing that the Internet is a field of communications that the Supreme Court has firmly declared to be entitled to the highest level of First Amendment protection.[63]

The source of the confusion over the liability of ISPs as publishers is the Stratton Oakmont decision that clumsily applied a fault standard best suited for newspapers and book publishers upon Prodigy, merely because Prodigy attempted to provide a small measure of order and control over the content of its electronic bulletin boards. The amount of control Prodigy wielded is crucial to determine whether it belongs to the "publisher" subset. Such entities are held to "increased liability" because they are not "passive receptacle[s] or conduit[s]" as they exercise "editorial control and judgment" over the content they publish.[64] If we are to view an ISP as an entity that merely provides a means through which users can transmit written messages to each other then it should not be liable for "the defamatory character of [a] message unless [it] knows or has reason to know that the message is libelous."[65] In order to see why Prodigy's role in providing a public forum for the posting of material on its bulletin boards best fits the "distributor" subset, it is first necessary to take a more general look at society's interest in preserving the tort of defamation and the rationale for maintaining a publisher/distributor distinction within it.

Defamation is said to be a relational interest as it involves the "opinion which others in the community may have, or tend to have, of the plaintiff."[66] While a direct insult might sustain a claim of intentional infliction of emotional distress, it will not support a defamation claim if that communication is not conveyed to a third party.[67] The derogatory words must "diminish the esteem, respect, good-will, or confidence in which the plaintiff is held, or . . . excite adversary, derogatory or unpleasant opinions against him."[68] It is not enough that the plaintiff personally feels the derogatory words have diminished his standing in the community, they must be the kind of words that would either be defamatory in virtually any case (for example, "an anarchist . . . eunuch . . . immoral . . . coward . . . drunkard"),[69] or given the current circumstances and activities of the defamed (for example, saying that an incumbent of a public office requiring citizenship is not a citizen, or a kosher meat dealer is accused of selling bacon).[70] In addition, it is not necessary that anyone believe the allegation to be true since the fact that the words have circulated at all must be to some extent injurious to the plaintiff.[71] Only a defamatory meaning must be conveyed;[72] however, it is crucial that the actual statement be false, as well as defamatory.[73]

Publication is integral to defamation liability.[74] It is at this point that the publisher/distributor distinction becomes crucial to the tort for both practical and policy reasons. A publisher, like a newspaper or publishing house, is held to the higher standard because it has "the opportunity to know the content of the material being published and should therefore be subject to the same liability rules as are the author and originator of the written material."[75] For this reason it is subject to liability without proof of its actual knowledge of the content's possible defamatory nature prior to its publication, because it is deemed to have been negligent in failing to ascertain its falsity or defamatory character.[76]

Mere distributors, on the other hand, are only liable for defamation if they know or have reason to know of the defamatory article.[77] Because society values freedom of expression and wide public access to news and literature, we do not wish to discourage vendors from stocking materials for fear of liability. Therefore, if there are no facts or circumstances evident to the disseminator that would indicate to a reasonable person, upon inspection, that a particular book, magazine, or newspaper contains defamatory content the distributor will not be found liable for defamation.[78] A distributor is not required to examine the material it disseminates and, absent special circumstances that would warn it in advance that the content is defamatory, it "is under no duty to ascertain its innocent or defamatory character."[79]

Though there is a place for some publisher liability for ISPs over the Internet, it should only be in cases closer to Drudge where the ISP takes on a publisher's role. The Stratton Oakmont court should not have associated the moderation of newsgroups with that role.[80] The moderators of those forums did not solicit articles or pay the participants as a true editor would have done.[81] They simply organized the postings and regulated content in the same way that the moderator of a town hall meeting would have.[82] Because such a moderator would not be held liable for defamation unless she knew in advance that she was facilitating defamatory statements, neither should Prodigy have been found liable. Though the Board Leaders had the ability to monitor incoming transmissions and to delete offensive content, that does not make them an editorial staff. Bookstores and libraries make similar decisions in judging what material to make available and are allowed to take into account its potentially offensive nature without incurring the status of "publishers" for defamation liability. Though bookstores might not use software to screen offensive language, ISPs must deal with the dilemma of handling thousands of mostly anonymous postings in a fair and efficient manner, while at the same time making its boards appealing to those who want to exchange on-topic information without having to wade through numerous infantile postings full of profanity.

ISPs serve an important role in providing such outlets for the public to exchange opinions and information. If an ISP wishes to limit all or some of its boards to notes which are not in "bad taste or grossly repugnant to community standards"[83] and are conducive to a "harmonious online community,"[84] it should be able to do so without incurring the liability of a traditional publisher. If an individual feels that one board's content restrictions are too strict he can always move to a board that is not as restrictive or to one of the thousands of public newsgroups that do not censor content.[85] Unlike the assertions made in a newspaper or magazine, which have an extra veneer of truth because the public knows the articles are approved and commissioned by editors, the public likely holds no similar illusions about the postings in a public forum. The guidelines for the Prodigy "Money Talks" board specifically stated that Prodigy was "committed to open debate and discussion."[86] No assertions were made about the reliability of the postings. The control exercised was merely used to sustain the harmonious nature of the board.[87]

If an ISP is liable "only" under the distributor standard it still must deal with the problem of how to react once it has been put on notice that an assertion by a third party broadcast from one of its servers is possibly defamatory. If the ISP were simply to remove every bulletin board, newsgroup, or web site posting once one person complained, it could create a chilling effect on Internet speech. The Supreme Court has recognized that unjustified fears of liability for libel can produce a chilling effect on speech that is antithetical to the First Amendment.[88] In addition, the Zeran court partially justified its decision by finding that liability on notice has a "chilling effect on the freedom of Internet speech."[89]

When private citizens have the power to abridge free speech in a way that the government does not it creates a power that has been referred to as the "heckler's veto."[90] The term was coined to describe instances where "the state suppresses the free protest of unpopular groups because their protest causes observers to want to do them harm."[91] It has also been used to describe any occasion where a speaker is "heckled down" by an audience member.[92] Thus, it may seem easier to allow the ISP to provide a forum for unregulated speech without consequences. However, the conflict between the right of the individual to speak and the right of the individual to preserve his good name has always required a difficult and often complex balancing of these competing interests. Up to this point, however, no other media has been able to exempt itself from liability for this centuries-old tort.[93] If there is to be a way around the "heckler's veto," however, we first must confront one of the Internet's primary features: anonymity.

If a person feels she has been defamed through an assertion made on the Internet, one great stumbling block she might face is the ease with which the medium facilitates anonymity. For example, the plaintiff in Zeran v. America Online was left without recourse once the court held AOL to be immune from liability as a distributor of third party information content because the messages had been posted by an anonymous person whose identity was never able to be traced.[94] Since Zeran was not an AOL subscriber he had no access to AOL's content areas, while those who did subscribe had the opportunity to damage him by defaming his reputation at will in a "libel-free" zone in which the authors are virtually free from liability due to anonymity and the ISP is free due to its immunity from being considered liable as a distributor or a publisher.

Anonymity has been both one of the most praised and vilified features of the Internet. While AOL and other ISPs with subscriber-only chat rooms and bulletin boards allow users anonymity through the use of user names, anyone can take advantage of an even harder to trace method that disguises the identity of her messages-anonymous remailers. Such services are widely available to anyone who has access to the World Wide Web.[95] Anonymous remailers work by sending a message along a chain of remailing programs that use cryptography to continually forward the communication in ways that disguise the identity of the original sender.[96] These remailers act as a cryptographic equivalent of the CDA "Good Samaritan" provisions for private figures by allowing "users to avoid being held responsible for the content of the messages they send."[97] Ironically, the remailers also disguise the user's Internet service provider.[98]

This anonymity can be one of Cyberspace's greatest virtues by allowing people the freedom to post in newsgroups and bulletin boards without fear of being harassed or unfairly criticized for their views. One federal court has noted that proponents of anonymity feel that it protects them "on a regular basis for the purpose of communicating about sensitive topics without subjecting themselves to ostracism or embarrassment."[99] However, anonymity can also protect those who wish to damage reputations and frustrate efforts by an ISP to determine whether a complained-about-posting should be removed. If the poster is anonymous the "heckler's veto" becomes even more powerful since the free speech concerns of an unknown person are much more likely to be usurped by the claims of an identifiable potential litigant.

For those who wish to continue the current regime of non-publisher status for ISPs, a state or federal law restricting the use of anonymous speech might seem a more viable solution. After all, it would target the creators of the speech instead of those who merely provide access to it. However, the Supreme Court has already shown that it will subject content-based restrictions on free speech over the Internet to its strictest scrutiny.[100] In addition, a federal court has already struck down a Georgia law that attempted to impose criminal liability on those who falsify their identity over the Internet.[101] The court specifically cited the statute's lack of a requirement of an intent to deceive as evidence that the statute was not narrowly tailored enough to present a compelling state interest to restrict the content of speech.[102] Since such an intent would be extremely difficult to prove anyway, a state or federal statute banning anonymous speech would seem to have little chance of succeeding as a deterrent to defamation. In addition, ISPs would act as true "distributors" in this context, and should only be found liable if they had notice and were given a reasonable amount of time to remove the anonymous defamatory posting.

Another potential method to lessen the "chilling effect" of notice-based liability would be for Congress to adopt a notification process for defamation similar to the one used for potential copyright violations in the Digital Millenium Copyright Act.[103] Under the Act a written notification of claimed copyright infringement must be sent to an agent designated by the ISP and must include a name to contact, specific identification of the copyrighted work, the address on the Web where the material is located, a good faith statement that the material is copyrighted and that the use of it by the party in question has not been authorized, and a statement that the information is accurate and that the complaining party (under penalty of perjury) is authorized to act on behalf of the owner of the right that has been allegedly infringed.[104] If the communication does not specifically comply with these requirements then it cannot be used in considering whether the service provider has been given "actual notice or is aware of facts or circumstances from which infringing activity is apparent."[105] If the notice does comply with the Act then the ISP must respond expeditiously to remove or disable access to the material or risk liability.[106]

Such a procedure could work effectively in the context of distributor liability for defamatory statements as well. A set procedure would remove some of the uncertainties about liability and lessen the chilling effect of the "heckler's veto." Each complainer could be required to swear (under penalty of perjury) that she has made her accusation of libel in good faith, and to make a detailed factual accounting of her basis for claiming that the assertion is false and injurious to her reputation. If the accusation did not provide the ISP with the basis to determine if the statement was defamatory, then the ISP could be deemed to have not received sufficient notice for notice-based liability.

In a factual scenario like that in Zeran v. America Online, the process would be simple. Once Zeran became aware of the posting that claimed to offer for sale T-shirts mocking the Oklahoma City bombing and featuring his phone number, he could send a written communication to the ISP's designated agent swearing that he offered no such product. He could also detail the numerous crank calls he had received and the damage it had done to his business. Armed with this notice the ISP could feel justified in removing the posting, especially considering that the poster was untraceable. A requirement of providing detailed information concerning both why the assertion is false and the specific damage it had caused would likely discourage many of those who might wish to exercise a "heckler's veto" for ulterior motives. It also provides the individual ISP with an industry-wide standard such that it does not have to fear that it will lose business to other ISPs that exercise no control whatsoever. Users will likely have greater respect for a process that only requires that ISPs remove posts upon receiving detailed notice than one that works according to the dictates of the "heckler's veto."

In addition, the non-cyberspace related case of Tacket v. General Motors Corp.[107] shows a way that ISPs could survive a notice-based regime of liability just like all other distributors, even without the aid of a Federal mandated notice process. That case involved an employee's claim that he had been defamed and that his injuries had been aggravated because his employer had intentionally and unreasonably failed to remove a sign that had been posted in the plant where he worked. The employee claimed that the sign disparaged his reputation by accusing him of being a thief.[108] The sign was allowed to stand for six or seven months before it was finally painted over by plant employees.[109] The court held that since "a reasonable person could conclude that Delco 'intentionally and unreasonably fail[ed] to remove' the sign" it had therefore published its contents.[110] In examining a series of cases involving the question of whether the gap between removal and notification was significant enough to constitute negligence, the court proposed a simple cost-benefit test: if "the costs of vigilance are small . . . and the benefits potentially large" then negligence can be inferred.[111] Such a test could be applied to Cyberspace if courts took into account the amount of time it took a reasonable ISP to investigate a defamation complaint and remove it. Relevant factors could include the size of the ISP, the nature of the posting, the potential harm, the location of the posting, and the popularity of the site. Due consideration of these factors could provide a workable standard that both recognizes the unique attributes of this relatively new medium, and provides a remedy to those who reputations are defamed. Otherwise one could find oneself in a cyber-equivalent of the plant in the Tacket case.

None of these measures is perfect, of course. It is much easier for ISPs to judge whether a copyrighted work has been violated since the distribution of an album track or software without permission is much more obvious on its face than whether a factual assertion is defamatory. It is certainly true that ISPs post thousands of user messages a day, and the decision-making process as to whether they should remove a posting upon notice would not be an easy one. However, these concerns beg the question of whether we should have liability for third party content for all members of the distributor subset, not just ISPs. While it is true that ISPs have no special expertise to apply in judging whether assertions are false, neither do news dealers and book stores.[112] Beyond merely pointing to the CDA and saying, "because Congress said so," it is hard to ferret out a policy rationale that justifies allowing AOL to tout itself as a service that offers to its readers information from "a gossip columnist or rumor monger"[113] without fear of incurring liability (even upon notice), but allows a book store which sells a newspaper or book containing identical pieces written by the same author to be found liable if it had sufficient notice because it offered for sale a work that "notoriously persist[ed] in printing scandalous items."[114]

The Zeran court's holding that distributor liability for ISPs would create "a natural incentive simply to remove messages upon notification" that could result in a "chilling effect on the freedom of Internet speech,"[115] fails to take into account that the indiscriminate removal of internet web sites and bulletin board postings would alienate the ISPs' subscribers. In addition, many service providers regulate themselves anyway, because it pays to do so financially. For example, AOL censors its content itself and provides additional software to parents to censor content to their own specifications.[116] This was indeed a concern of Congress and the courts in the time between the ruling in Stratton Oakmont and the enactment of the CDA. The Zeran Court could have eliminated the danger of ISPs being hurt for taking a hands-on approach to editing content if they had applied the most common-sense interpretation of the CDA-that ISPs are not liable as publishers for third party information content, but can be as distributors.

It seems only fair that ISPs, when they act as distributors of third party content, should operate under the same rule as other providers of information from third parties. If there is a consensus that notice-based liability is unfair to distributors in a complex modern economy and digital age then all should receive the benefit of that determination, not just ISPs. We can, however, reach a common ground. As some commentators have already suggested, common law distributor liability already provides strong protection to ISPs. It is far from clear that ISPs need substantial protection against private individuals, who will almost surely be underfunded in comparison.[117] It is hard to justify the current regime when one considers the fact that a large corporation like AOL is immunized from liability for distributing defamatory materials while the neighborhood bookstore is not. ISPs have become one of our greatest distributors of speech, and it seems both inefficient and unjust to hold them to a lower standard of liability for defamation than all other existing distributors.

Even if an ISP were held to be a member of the publisher subset because it acted in an editor's role in regard to a news service like the "Drudge Report," it would still have significant protection. As a provider of its own news content the ISP would enjoy the protections of the same "actual malice" standard available to all citizens when commenting upon public officials.[118] In New York Times v. Sullivan[119] the Supreme Court set aside the common law strict liability regime for defamation that essentially found that the media published at its own peril.[120] In Sullivan, the Court held that a state court regime that allowed "a good-faith critic of government [to be] penalized for his criticism . . . strikes at the very center of the constitutionally protected area of free expression."[121] Thus, a public official who is defamed in regard "to his conduct, fitness or role as public official" cannot bring a defamation action unless he can prove the defendant knew of "the falsity of the communication or acted in reckless disregard of its truth or falsity."[122] Thus, even if ISPs were held to be within the publisher subset when they commission and pay for news articles, in many cases their liability would be notice-based anyway because many articles would involve public figures and officials and would therefore trigger the actual malice standard.

It is this actual malice test for defamation of public officials or public figures that Drudge plaintiff Sidney Blumenthal would almost surely have had to have overcome if the Court had not ruled that the existence of the CDA precluded the question of liability.[123] He would not have had to prove genuine ill will toward himself by AOL, but would have had to show by clear and convincing evidence that AOL knew or had reckless disregard for the fact that the assertions about Blumenthal were libelous.[124] "The standard of actual malice is a daunting one,"[125] and most cases using the actual malice standard are resolved in favor of the defendant. One could take the position that the Drudge decision was no great loss since Blumenthal, because being an advisor to President Clinton and a well-known writer, he likely would be considered both a public figure and a public official.

The trial forum, however, is crucial in such cases because the actual malice standard requires a "subjective" analysis where the court looks to the specific "state of mind" of each defendant.[126] It will not be enough for the plaintiff to prove that the publisher failed to investigate the report.[127] The publisher "must act with a 'high degree of awareness of . . . probable falsity.' "[128] Only when the publisher is aware that the veracity of the source is in doubt can it found liable for a reckless disregard of evidence-if it fails to look at evidence which is within easy reach.[129] The evidence at trial will have to be strong enough to support a finding of "highly unreasonable conduct constituting an extreme departure from the standards of investigation and reporting ordinarily adhered to by responsible publishers."[130]

The above standard could certainly be adapted to address the problems which are unique to Internet journalism. As an ISP publisher would likely be less hands-on than a magazine editor, the time it takes a responsible Internet publisher to investigate allegations of defamation might reasonably be longer. As applied to Drudge, what would be crucial is determining what knowledge Drudge had of the "top GOP operatives"[131] who were the sources for his story. Drudge implied later on that he had essentially been tricked by a partisan source, saying "I've been had."[132] If a court found that Drudge based his report on an unverified anonymous phone call or had obvious reasons to doubt the source's veracity, he could still be found liable under the actual malice standard.[133] Even if it could have been found that Drudge acted with reckless disregard for the truth, however, such a finding would not prevent a factfinder from absolving AOL even if was held to belong to the "publisher" subset.

Even putting aside the issue of AOL's separate state of mind, it would stand a good chance of avoiding liability for arrangements like the one it had with Matt Drudge on respondeat superior principles. While it is true that in some circumstances an employee's negligence or actual malice has been imputed to its publisher,[134] if the writer is found to be an independent contractor then the writer's state of mind cannot be attributed to the employer.[135] In one federal court decision, the court found an independent contractor relationship between The New Yorker and one of its article writers (Janet Malcolm) despite the fact that the she had a long-term relationship with the magazine and an office at their headquarters.[136] It found that "if control may be exercised only as to the result of the work and not the means by which it is accomplished, an independent contractor relationship is established."[137] This is exactly the kind of control AOL exercised over Drudge. AOL reserved only the right to "remove content that AOL reasonably determine[d] to violate AOL's then standard terms of service."[138] Both writers were paid a yearly salary, but Drudge did not maintain an office at AOL or have a relationship with it prior to the agreement as Malcolm did.[139] In addition, the New Yorker's editing was much more extensive than AOL's.[140]

Thus, one can see that it is not necessary to absolve ISPs of publisher status to protect them from unjust liability. If the courts found ISP/news-service writer relationships to be more akin to employer/independent contractor relationships than an employer/employee one, ISPs would not have to worry about having the writer's liability imputed to them regardless of whether their role was more suited to a publisher or distributor status. One problem AOL would encounter in making such a claim in regard to Drudge is the fact that it touted him as a rumormonger and one who could supply "instant gossip" to its subscribers.[141] The plaintiff might claim that AOL has no right to claim it relied on Drudge's reputation as a reporter if it was instead relying on him to provide rumors and gossip. These are certainly difficult questions about the role of an ISP as publisher, but they will never have a chance to be answered unless the courts are willing to hear claims that ISPs are responsible as publishers if they act like publishers.

Many commentators welcome the current liability regime for ISPs, and some have already proposed that any party who participates over the Internet should be deemed a "public figure" for purposes of determining liability for defamation committed in Cyberspace.[142] These arguments stem from the Supreme Court's ruling in Gertz v. Robert Welch, Inc,[143] in which the Court designated two bases for determining whether an individual is a public or private figure. The first is a general one in which an individual has achieved "such persvasive fame or notoriety that he [has become] a public figure for all purposes and in all contexts."[144] The second is that of a limited public figure who "voluntarily injects himself or is drawn into a particular public controversy and thereby becomes a public figure for a limited range of issues."[145] Those who advocate that all Internet participants are public figures look to the policy rationales for both of these types of public figures in justifying their position. Like general public figures, Internet participants are said to be worthy of a lesser protection against libel because they have access to the same medium to defend themselves against defamatory statements. In addition, like limited public figures, they have injected themselves voluntarily into the public controversy through the same medium in which they claim to be defamed.[146]

This position, however, exaggerates the equality of position of the defamer and the defamed. First, not all Internet authors are equal. The author of a large public web site like the Drudge Report which registers millions[147] of visits each month cannot be said to have the same access to the media as an ordinary Internet user who has a small local web site or posts to newsgroups or bulletin boards. It would be equivalent to requiring that the host of a public access show in Ann Arbor, MI who brought suit in Michigan, prove actual malice to prove that Jay Leno was liable for a defamatory statement because they both had access to the same medium and locals could just as easily access both shows. No one would find the private individual's show because no one would be looking for it-the same as a web site. As far as they are both interjecting themselves into the same controversy, it is hard to say this is true if the public only sees one side of the controversy. This is not a problem with criticism of public figures because the media usually will broadcast their opinions and they will reach a stage equal to the defamation. Plus, the same people interested in the defamation will be more likely to follow the retraction in the case of a public figure. The Supreme Court addressed these concerns in Gertz:

More important than the likelihood that private Individuals will lack effective opportunities for rebuttal, there is a compelling normative consideration underlying the distinction between public and private defamation plaintiffs. An individual who decides to seek governmental office must accept certain necessary consequences of that involvement in public affairs. He runs the risk of closer public scrutiny than might otherwise be the case . . . . Those classed as public figures stand in a similar position . . . . [They] have thrust themselves to the forefront of particular public controversies in order to influence the resolution of the issues involved. In either event, they invite attention and comment.[148]

AOL-by adding Drudge's material to its subscriber-only area, paying him, becoming his sole source of income, and reserving the right to edit his content-acted similarly to a publisher. It should be not be able to avoid liability merely because its editorial content travels to users through a network of computers. A compromise position of only allowing ISPs to have publisher liability when they assume a publisher's role would both address the complexities of the new medium while at the same time avoiding the costs of a regime which provides little remedy to private individuals who are defamed. There are already significant protections within the law that would protect ISPs from unjustly bearing responsibility whether they acted as publishers or distributors. Any solution to this situation will be difficult, however, due to the jurisdictional problems inherent in the Internet and the federal courts' current interpretation of the CDA. In light of these two factors, the only viable solution that will be fair to both ISPs and future victims of libel is a federal standard.

The jurisdictional issues related to the Internet are unique, and lend themselves to the more all-encompassing utility of a federal standard. No one is really sure how large the Internet really is, and approximately 3/5ths of the 9 million host computers that exist world wide are located in the United States.[149] The Internet is largely ungoverned and is slightly chaotic due to its freeform nature. Whereas the jurisdiction of states is based on geographical boundaries, cyberspace flows through, across, and outside of the United States oblivious to those boundaries.[150] As one court has noted, it is the "unique nature of the Internet [that] highlights the likelihood that a single actor might be subject to haphazard, uncoordinated, and even outright inconsistent regulation by states that the actor never intended to reach and possibly was unaware were being accessed."[151] A state that attempts to legislate the liability of ISPs could run afoul of the dormant Commerce Clause. It is on this basis that, in American Libraries Association v. Pataki, a U.S. District Court struck down a New York criminal statute that attempted to make it a felony to intentionally use the Internet to initiate or engage a communication with a minor that contains sexual content.[152] The Court stated that since the practical impact of the law would be an "extraterritorial application of New York law to transactions involving citizens of other states" the law was "per se violative of the Commerce Clause."[153] In advocating a federal standard for Internet law the Court went on to say that "the unique nature of cyberspace necessitates uniform national treatment and bars the states from enacting inconsistent regulatory schemes."[154]

Of course, ISPs when acting in a capacity as "publishers" of defamatory material in America already reap the benefits of a federal standard. Unfortunately, the standard is no liability. Due to the Internet's non-jurisdictional nature and its ability to provide anonymity to all participants, the current standard leaves little remedy to those individuals whose reputations have been damaged through the use of Internet communication. Defamation law has traditionally been left to the states, and the Supreme Court has held that they may "define for themselves the appropriate standard of liability for a publisher or broadcaster of defamatory falsehood injurious to a private individual"[155] as long as they do not impose liability without fault.[156] Though states have a legitimate interest "in compensating private individuals for wrongful injury to reputation,"[157] a federal standard imposing some liability is now necessary because we have a federal, largely court-imposed, standard for ISPs barring virtually all liability.

Since authors of defamatory statements are harder to identify on the Internet and the scope of the medium is worldwide, it is unfair to hold carriers who unwittingly disseminate such messages over the Internet to a lower standard of responsibility than all other distributors. Congress needs to create a new federal standard to correct the ambiguities present in the "good samaritan" provisions of the Communications Decency Act, as the Courts have misinterpreted the provision to give immunity from defamation suits to ISPs for information content that they do not create themselves. Such reform seems particularly apropos considering that the major provisions of the Act were overturned by the Supreme Court in Reno v. ACLU. Since the courts' current interpretation of the "good samaritan" provisions of the CDA provides less protection than is needed for libel victims and more protection than is needed for ISPs, Congress should amend those provisions to allow an ISP to be held to the same liability standard as the media it most closely resembles given the nature of the ISP and the role it plays as content provider in the particular case at issue.

If Congress were to clearly define the meaning of the word "publisher" in the CDA to make it clear that distributor liability for third party content is possible, that would certainly be an improvement. But that interpretation would also fail to take into account the growing ways in which ISPs act like members of the "publisher" subset. Amongst modern communications mediums, ISPs are unique in the ways they can act as both publishers and distributors of written content. In our efforts to protect them from unjust liability for the former, we should not preclude liability for the latter if it is deserved. In order to prevent a federal standard for liability from having a "chilling effect" on free speech, Congress should also provide a standard for the notification process similar to that used in the Digital Millenium Copyright Act. Such a process is necessary because it is impossible to both allow completely unfettered free speech on the Internet and still protect the interest an individual has in her good name and reputation. If Congress were to reform the CDA and pass legislation stating that ISPs should be considered in most instances a distributor but in certain rare cases a publisher, it would create a liability standard that would do the least harm to the forum for free speech over the Internet. Such a standard would stand the best chance of preserving the right of individuals to freely speak their mind and communicate with each other in Cyberspace, while reviving the tort the common law developed to protect the interest each person has in preserving his good name that makes others willing to listen.