Richmond Journa

l of Law &Technology

Volume VI, Issue 2, Fall 1999


A Unified Approach to Cyber-Libel:
Defamation on the Internet, a Suggested Approach

Barry J. Waldman [*]


Cite As: Barry J. Waldman, Comment, A Unified Approach to Cyber-Libel: Defamation on the Internet, a Suggested Approach, 6 RICH. J.L. & TECH. 9 (Fall 1999) <>.[**]

Table of Contents:



A. General Background

B. Some General Approaches To Online Personal Jurisdiction

1. Totality of Contacts Approach

2. Effects Test Approach

3. Keeton Test Approach

C. Suggested Approach: Nature of the Website, Zippo Test


A. General Background In Defamation Cases

1. General Tort Approach

2. Traditional Multi-State Defamation Approach

B. Suggested Approach: Focus On Domicile and Location of Greatest Effects

1. General Tort Approach

2. Traditional Multi-State Defamation Approach


A. General Background

B. Historical Cyber-Libel Development

1. Cubby, Inc. v. CompuServe, Inc.

2. Stratton Oakmont, Inc. v. Prodigy Services Co.

3. Communications Decency Act of 1996 § 230(c)

C. Current Approach

1. Zeran v. America Online, Inc.

2. Blumenthal v. Drudge

D. Suggested Approach: Original Author Liability, Substantive Editor - Service Provider Liability


A. Common Law

B. Current Approach/Statutory





{1} The Internet is a global "super-network of over 15,000 computer networks used by millions of individuals, organizations, corporations and educational entities the world over."[1] As the Internet has developed, it has become a medium not only for entertainment, but also an important source of information and news distribution.[2] Because the Internet and the World Wide Web[3] have developed into important resources for information and news, traditional media concerns and legal controversies have reached a level of growing importance. While it is true that much of the "development of the law concerning the . . . Internet [i]s in its infant stages,"[4] online defamation has received significant comment.[5]

{2} This paper will differ from the traditional means of analyzing Cyber-Libel,[6] by providing a unified approach to the determination of rights and recoveries by the parties. It addresses the general and computer law approaches, and provides a suggested Cyber-Libel approach to the issues of jurisdiction, choice of law, and defamation law.[7] This survey is intended to provide a simple, coherent means to dealing with Cyber-Libel, which if employed, would result in consistently fair results for both the plaintiff and defendant.




A. General Background

{3} Due process requires that a non-resident defendant stand trial in a particular court only if the court can properly exercise personal jurisdiction over the defendant.[8] In determining whether or not jurisdiction lies, once contested by the defendant, it is the duty of the plaintiff to establish a "prima facie showing that jurisdiction exists."[9] The allegations relating to the propriety of jurisdiction are presumed true to the extent that they are not in conflict with the defendant's affidavits. Once the prima facie case is put forth by the plaintiff, "all factual disputes [relating to the exercise of jurisdiction] are resolved in favor of the plaintiff." [10]

{4} After the facts underlying the purported jurisdiction have been found, they must be applied to the jurisdictional maxims of the state in which the court sits.[11] Many states have opted to extend jurisdiction to the limits of due process; other states have chosen to particularly define the grant of jurisdiction. In either scenario, the court must, after determining if jurisdiction exists under the state's jurisdiction statutes, inquire whether or not the exercise of jurisdiction violates constitutional due process.[12] To satisfy the due process requirement, the "assertion of personal jurisdiction over a nonresident defendant [must] be predicated on 'minimum contacts' between the defendant and the [forum] State."[13] In assessing the quality and nature of contacts required for jurisdiction, the court determines if the exercise of personal jurisdiction is based upon either general jurisdiction or specific jurisdiction. In making this distinction, "[s]eparate tests have developed to determine whether a forum has general or specific jurisdiction over a defendant."[14]

{5} General jurisdiction is the court's exercise of personal jurisdiction over a defendant based upon his contacts with the forum, which are unrelated to the cause of action plead by the plaintiff. It requires that the defendant be engaged in "systematic and continuous" activities with the forum state.[15] The exercise of general personal jurisdiction over a non-resident entity presents a difficult burden and is often rejected as a basis for the exercise of jurisdiction by courts.[16]

{6} Specific jurisdiction lies where the contacts relied upon to establish the court's personal jurisdiction over a defendant include the activity complained of by the plaintiff.[17] As at least one circuit has defined the test for specific jurisdiction as requiring:

1) the nonresident defendant must do some act or consummate some transaction with the forum or perform some act by which he purposely avails himself of conducting activities in the forum, thereby invoking the benefits and protections [thereof];

2) the claim must be one which arises out of or results from the defendant['s] forum related activities; and

3) [the] exercise of the jurisdiction must be reasonable.[18]

{7} This test essentially requires that the court find contacts, "such that maintenance of the suit does not offend traditional notions of fair play and substantial justice."[19] Often a court focuses on whether or not the defendant, through his actions, could have anticipated being hailed into court in the forum state.[20] If the court is able to make this finding, jurisdiction will likely be upheld.

{8} Determining personal jurisdiction in a Cyber-Libel case can be a difficult proposition. Unlike the traditional forms of real-world tortious action, Cyber-Libel does not have a specific locus of bad act. The potential anonymity of the tortfeasor, [21] and the potentially-wide distribution of the injury causing material may make identification of proper jurisdiction unusually difficult. It is not unheard of for mere children to have the capacity to inflict great damage technologically on a business or the government, with but a few years of familiarity with computer use. If an individual needs such limited training to become a "hacker,"[22] it is easy to see how others with even less experience can become a tortious threat on the Internet. With this ability to hide true identity, and the wide-dispersing effect of the technology, jurisdictional issues can be crucial. Thus, a clear and reliable treatment of this issue is crucial to consistency.

B. Some General Approaches To Online Personal Jurisdiction

1. Totality of Contacts Approach

{9} One approach taken by the courts to determine jurisdiction for actions based, at least in part on Internet contacts, is a totality of contacts analysis. Using this method, the court considers the online contacts with the forum and the more traditional forms of contact.[23] The court gathers and weighs all contacts that the defendant has with the forum, either electronic or non-electronic, to determine whether sufficient minimal contacts exist to assert specific jurisdiction.[24] In EDIAS Software International, L.L.C. v. Basis International Ltd.,[25] the court found jurisdiction by looking to both Internet and non-Internet based contacts. The court held that Basis purposely availed itself of the Arizona forum by posting defamatory statements to its website and by e-mailing specific individuals in the Arizona forum these same defamatory declarations. The court also, however, relied upon the facts that Basis entered into a contract with an Arizona corporation, contacted EDIAS by telephone in Arizona, and sent facsimiles, invoices, products, and employees to Arizona.[26] The court seemed to look at the totality of these contacts to determine that Basis had availed itself of the Arizona forum.[27]

{10} Similarly, in Rubbercraft Corp. of California v. Rubbercraft, Inc.,[28] the United States District Court for the Central District of California took a totality of contacts approach. After explaining that a passive website alone is insufficient to convey jurisdiction in the forum, the court proceeded to evaluate other non-Internet contacts to determine whether jurisdiction should lie.[29] The court identified Rubbercraft's contacts as advertisement by webpage and a nationally-circulated periodical, maintenance of an "800 number phone line," operation of a webpage and sales in the forum of about $20,000 in 1997.[30] The court employed a totality approach explaining that, "[a]lthough each of these factors may not alone create purposeful availment [and the other elements required for jurisdiction], in combination they satisfy the ... test of affirmative conduct ..." sufficiently to establish specific jurisdiction.[31]) In Cybersell, Inc. v. Cybersell, Inc.,[32] the United States Court of Appeals for the Ninth Circuit took a nearly identical approach, in that it found that a passive website was insufficient to establish jurisdiction, and subsequently employed a totality of contacts analysis.[33]

2. Effects Test Approach

{11} Another popular approach taken by courts to analyze online jurisdiction is to employ an effects test. This approach essentially inquires whether the "forum state 'is the focal point of the [defamatory] story and the harm suffered.'"[34] The inquiry requires an examination of whether the defaming party had the intent to cause injury in the forum or whether knowledge could be reasonably presumed that injury would be caused in the forum.

{12} The court employed this effects test in the EDIAS case.[35] In this case, the plaintiff argued for the exercise of jurisdiction through an effects test analysis, contending that the defendants, through their Internet actions, knew or should have known that the statements made would create a harm in the forum.[36] The court agreed with this approach and found that, "[b]asis directed the e-mail, [w]eb[]page and forum message at Arizona because Arizona is EDIAS' principle place of business. EDIAS allegedly felt the economic effects of the defamatory statements in Arizona."[37] The court, relying upon California Software, Inc. v. Reliability Research,[38] explained that, just because the defendant did not send a tangible object into the forum or was not present in the forum, does not mean that it should be able to use the unique characteristics of Internet technology to avoid jurisdiction.[39] Therefore, the court found the effects test to be an appropriate rule for determining jurisdiction over the defendant.

3. Keeton Test Approach

{13} The Keeton test approach to jurisdiction transports a traditional multi-state defamation jurisdictional analysis into Cyber-Libel. This approach expands on the Single Publication Rule of defamation and allows for the exercise of jurisdiction where the defamatory statement is published and delivered to the forum.[40] The Supreme Court explained that the defendant's "regular circulation of magazines in the forum State is sufficient to support an assertion of jurisdiction in a libel action based on the contents of the magazine [circulated in the forum]."[41] The Court went on to explain that "regular monthly sales of thousands of magazines cannot by any stretch of the imagination be characterized as random, isolated, or fortuitous," and that "[t]he tort of libel is generally held to occur wherever the offending material is circulated."[42] This would hold true even if the defamed party were essentially anonymous in the forum prior to publication.[43] Thus, the circulation of even a single copy of the defamatory statement in a forum would give rise to a tort action, and is sufficient contact for the exercise of jurisdiction.

{14} Applying this approach to jurisdiction in Cyber-Libel provides a unique quandary because the defendant is likely to argue that he made no purposeful entrance into the state where a generally available website was accessed in the forum state. The defendant would probably argue that he did not have intention to "circulate" or "deliver" the statement into the forum. At least one court, however, has employed a Keeton-type approach to the posting of a defamatory statement by a non-resident defendant on a passive website.

{15} In TELCO Communications v. An Apple A Day,[44] the court addressed whether a defamatory statement on a passive website is circulation, delivery, and advertisement within the forum state. The court adopted the Keeton approach and found that, even though the evidence showed that the defendant had no other actual contacts with the forum, An Apple A Day was subject to the exercise of jurisdiction.[45] The court explained that defamation occurs wherever the material is "distributed," and that the defendant "should reasonably have known that the press releases would be received in Virginia."[46] Thus, while the defendant took no assertive action to ensure that the defamatory statement would be received in the forum, the knowledge that the statement would be circulated and delivered to the forum is sufficient to establish jurisdiction under the Keeton model. [47]

C. Suggested Approach: Nature of the Website, Zippo Test

{16} Jurisdictional questions in a unique media such as the Internet, require a moderately unique approach. Blindly applying traditional contacts analysis models to a media which does not typically leave physical evidence or record of the contacts made, presents difficulties in determining a defendant's amenability to suit in a particular forum.[48] Because such an approach often proves ineffective, a more reasonable procedure to determine jurisdiction in Cyber-Libel cases would identify the nature of the electronic communication, and imply the level of contact that this communication has with a particular forum. Since, under traditional defamation law, mere delivery of the defamatory statement to a particular forum would establish amenability to suit, this approach serves to fairly approximate the conventional burden on the defendant.[49] Following a treatment of jurisdiction that most closely approximates traditional principles in defamation law is most likely to provide a consistent and dependable result. This nature of the website approach is embodied in Zippo Manufacturing Company v. Zippo Dot Com.[50]

{17} The court in Zippo established a sliding scale approach to the determination of jurisdiction over a non-resident defendant based on Internet contacts. The court explained:

[T]he likelihood that personal jurisdiction can be constitutionally exercised is directly proportionate to the nature and quality of commercial activity that an entity conducts over the Internet. This sliding scale is consistent with well developed personal jurisdiction principles. At one end of the spectrum are situations where a defendant clearly does business over the Internet. If the defendant enters into contracts with residents of a foreign jurisdiction that involve the knowing and repeated transmission of computer files over the Internet, personal jurisdiction is proper. At the opposite end are situations where a defendant has simply posted information on an Internet Web site which is accessible to users in foreign jurisdictions. A passive Web site that does little more than make information available to those who are interested in it is not grounds for the exercise [of] personal jurisdiction. The middle ground is occupied by interactive Web sites where a user can exchange information with the host computer. In these cases, the exercise of jurisdiction is determined by examining the level of interactivity . . . that occurs on the Web site.[51]

{18} This model for the exercise of personal jurisdiction provides protection from the unreasonable exercise of jurisdiction, while not wholly foreclosing suit against the libelant, even where the libelant merely "posts" the defamation to a passive website. [52] While it is true that when posting to a passive website, a defendant can reasonably assume that his words will be read in any number of jurisdictions, there are no reciprocal contacts with any forum in which he is not a resident.[53] The sliding scale, however, provides protection where the defendant specifically sends the defamatory material into a forum, such as with e-mail messages,[54] and protects where the website is produced by a traditional information source.[55]

{19} One area of controversy under the Zippo approach is the clarification of whether a passive website is sufficient to convey jurisdiction in a forum over a non-resident. While Zippo explicitly rejects passive websites as a basis for jurisdiction,[56] TELCO proceeded to analyze the passive website and exercise jurisdiction since it was, "advertising and soliciting over the Internet, which could be accessed by a Virginia resident [twenty-four] hours a day . . . [which] constitutes a persistent course of conduct ..."[57] This approach should be rejected under the suggested model, as it introduces liability for exposure to any individual or corporate entity in any forum which posts advertisements or information on the Internet. Such wide-ranging forum jurisdiction is manifestly a violation of the requirement that the exercise of jurisdiction be reasonable.[58] Furthermore, this approach subjects an individual or small company to jurisdictional liability in states and locations never imagined as a location of potential customers. The burden on the individual would thus be truly unreasonable.



{20} While the necessity for a uniform jurisdictional approach has been widely discussed,[60] a uniform approach to conflict of law issues in Cyber-Libel can prove to be just as important.[61] Unfortunately, the important issue of choice of law relating to Cyber-Libel has received little attention. The determination of which state law applies can determine such central issues as, what constitutes defamation, what level of proof is required, and what is the maximum amount of recovery?[62] While defamation law has its outer boundaries largely defined by principles of constitutional law, the variations in the state interpretation of defamation common law principles can be extensive. Just noting the variation in retraction statute protection, provides ample argument for the varying effect of a choice of law.[63] These differentiations in law are dramatic in the minds of the litigants, and therefore, are worthy of a separate analysis.

A. General Background In Defamation Cases[64]

1. General Tort Approach

{21} Choice of law questions in tort cases are generally controlled by section 145 of the Restatement on Conflicts.[65] This section requires an analysis of contacts with the interested states to determine which state's laws to apply. The inquiry is then guided by the general concepts of conflicts of law established in section 6 of the Restatement.[66] Section 145 of the Restatement makes the following inquiry: "(a) the place where injury occurred[;] (b) the place where the conduct causing the injury occurred[;] (c) the domicil, [sic] residence, nationality, place of incorporation and place of business of the parties[;] and (d) the place where the relationship, if any, between the parties is centered."[67] In this analysis, the court identifies the potential states whose laws may be applied.

{22} After determining which states have an interest in the application of its substantive law, the court applies the general principles of conflicts of law as identified in the Restatement's Section 6 to determine which of these states' laws are to be applied.[68] Section 6 of the Restatement directs an inquiry focusing on the following factors: which state has the greatest interest "in having its law applied[,] the relevant policies of the forum[,] certainty[,] [and] predictability [in the results reached by the state laws,] ease in the . . . application of the law to be applied[,] the promotion of interstate order[,] . . ." and the policy considerations underlying the substantive law to be applied.[69] These various factors are weighed to determine which substantive law should be applied.[70]

2. Traditional Multi-State Defamation Approach

{23} Section 150 of the Restatement on Conflicts of Law controls choice of law questions for the traditional defamation case. [71] The section applies to cases where the defamatory statement is contained "in any one edition of a book or newspaper, or any one broadcast over radio or television, exhibition of a motion picture, or similar aggregate communication ...."[72] The Restatement directs that the law of the state with the "most significant relationship" to the conduct giving rise to the cause of action be applied.[73] This determination is, like Restatement section 145,[74] guided by the principles established under section 6 of the Restatement.[75]

{24} The Restatement section 6 analysis is, however, supplemented by subsections (2) and (3) of section 150. Subsection (2) explains, "[w]hen a natural person claims that he [or she] has been defamed by an aggregate communication, the state of most significant relationship will usually be the state where the person was domiciled at the time, if the matter complained of was published in that state."[76] Similarly, subsection (3) provides for essentially the same domicile focus for corporate entities.[77] These subsections have, in many cases, the effect of eliminating the traditional Restatement section 6 analysis in favor of merely determining the domicile of the defamed party.[78] If, however, the plaintiff's reputation were not harmed in his state of domicile, the court would employ the Restatement Section 6 test.[79]

B. Suggested Approach: Focus On Domicile and Location of Greatest Effects

{25} While the traditional Multi-State Defamation approach may prove effective in handling conflicts of law issues, a more reasoned approach may be achieved. A more simplistic approach will ease the judicial burden of analyzing the complex contacts developed in Cyber-Libel. With the uncertainty created in anonymous Internet communications, and the complexity in developing a record of forum contacts,[80] a more direct approach may be needed. To be successful in easing judicial burden, this method must avoid making a contact-by-contact analysis of interested fora. The central goal in reaching a well-founded principle is to avoid over-complication, while reaching a reasoned legal approach.

{26} To achieve this goal, the most obvious and judicially sound point from which to begin is the domicile of the defamed party. In taking this tack, it is important to note that the analysis of Restatement Section 150 is an important guidepost.[81] The Wilson court explained, "[t]he state of a plaintiff's domicile is generally the place where most of his reputational contacts are found; therefore, 'the state of plaintiff's domicile generally has the greatest concern in vindicating plaintiff's good name ...' "[82] Thus, domicile should provide a generally accurate starting point for determining which state's law should apply.

{27} While it is true that the domicile state is most likely to be the most interested state in many defamation actions, situations will occur where the greatest effect of the defamatory statement is not felt in the plaintiff's domicile.[83] This problem can occur with frequency in the case of a corporate plaintiff because a corporate plaintiff may be a registered corporation in one state, yet conduct the vast majority of its business outside of that state.[84] Thus, for example, a film company formed as a Delaware corporation may suffer the greatest harm from defamation in California, where its potential employees reside. Such defamation and the resulting harms to the film company's reputation and goodwill may cause difficulty in obtaining big name actors and a proper technical crew. The same irrationality of choosing, de facto, the state of domicile in all cases may also arise in the case of an individual plaintiff.[85] It is for these very reasons that a secondary test must be developed.

{28} This secondary test should focus on the location where the greatest injury is caused by the defamatory statement. It is in this state that the plaintiff, who otherwise enjoys a positive reputation, loses the most value. The state has an interest in encouraging the maintenance of a positive reputation, as the individual or corporate entity will presumptively have a positive effect on some aspect of that forum. In order to determine the state where the greatest injury is incurred, an effects test seems well-suited and appropriately constructed to make an accurate determination. This test is a modification of the jurisdictional effects test, as applied by the court in EDIAS.[86]

{29} The central inquiry in the application of this secondary effects test is not whether the defendant knew that his actions would have an effect in the particular forum,[87] but instead, focuses upon the location of the greatest demonstrable impact of the defamatory statement. This impact can be shown through purposeful direction of the defendant's defamatory statement,[88] through actual shown economic harm, or through other relevant substantive and opinion evidence. The court then evaluates the evidence presented and makes a determination as to the state where the greatest impact is felt. It is important to note that this secondary effects test should only be employed when affirmatively plead by one of the parties, and should not be a sua sponte determination by the court. By placing this restriction on the secondary test, simplicity is encouraged, and the morass that choice of law determinations can occasionally produce is avoided.


A. General Background[89]

{30} Substantive law in the area of defamation is well defined. The following discussion is a basic review of the general principles of defamation as developed through common law principles. Much of what follows are the constitutional principles that set the outer limits of defamation liability as defined under First Amendment principles by federal courts. It is important to note that, while principles of constitutional defamation law seem well-grounded, to some extent, they are still new and experimental. In a greater scheme, constitutional defamation law is no more than a generation old[90] and is always subject to slow modification.

{31} Defamation is the "intentional false communication, either published or publically spoken, that injures another's reputation or good name."[91] In order to receive recovery in a defamation suit, the defamed party must first show that the defamatory statement was one of purported fact about the plaintiff.[92] To meet this burden, the plaintiff must demonstrate that a reasonable reader or recipient of the defamatory material would understand that the statement was in reference to the plaintiff. [93] The defamed party must then show that the defamatory statement was not subject to privilege,[94] and was "published" and "delivered" to a third party.[95] The plaintiff is then required to show that the defamatory statement caused some harm upon which recovery may be made.[96]

{32} The plaintiff's burden of proof is determined by assessing under which judicially established category of defamation suit his cause of action falls. These categories are established based upon the relative notoriety of the plaintiff. If the plaintiff is a pervasive public figure,[97] and the statement made is about a subject related to his fame, then the plaintiff must show that the defendant acted with actual malice.[98] Similarly, a public official criticized for his actions in the scope of his official duties must also prove actual malice by clear and convincing evidence on the part of the defendant.[99] If the plaintiff is a private citizen and the defamatory statement relates to a private concern, the plaintiff's burden is minimal because the statement is presumed false, and the defendant must demonstrate that the statement made was substantially true.[100] Finally, if the plaintiff is a private figure defamed about a subject of public concern, the defendant is subject to liability if the defendant knew the information to be false or acted in reckless disregard for the truth.[101]

{33} Additionally, the classification of the defendant will also affect the burdens established in a defamation case. A defendant may be either a publisher, distributor, or a common carrier.[102] A common carrier - an entity that has no editorial control over the information it carries - such as a telephone company, may not be held liable for information that it merely transmits from one party to another as a passive conduit.[103] A publisher, such as a newspaper or other entity that retains editorial control over the information it sends out, is held accountable if the traditional burdens discussed above are met, and at least negligence is shown in its actions.[104] A distributor, on the other hand, is often compared to a public library, and as such, may only be held liable upon a plaintiff's prima facie case and the showing that the distributor had actual knowledge of the defamatory content or should have reasonably known of the defamatory nature of the work.[105] Thus, a synthesis of the categorization of plaintiff and defendant reveals that, in a general sense, traditional defamation law imputes liability based upon the nature of the party defamed and the editorial control retained by the particular defendant.

B. Historical Cyber-Libel Development

{34} Cyber-Libel is an outgrowth of traditional defamation principles. As courts have struggled to adapt traditional print media applications to the electronic world, a number of legal fictions have been established and then eliminated. Judicial development has been supplemented with halting and stunted legislation. An effective approach must synthesize the positive attributes that have been historically developed with a forward-looking view of potential future issues. Additionally, any new suggested approach must address the limiting nature of congressional legislation, which has served as a positive influence, while at the same time, undercutting protections that lie at the heart of defamation law protections.

1. Cubby, Inc. v. CompuServe, Inc.

{35} In an early Cyber-Libel decision, the United States District Court for the Southern District of New York employed a publisher/distributor analysis to determine if an Internet service provider would be held liable for information available in its "electronic library."[106] CompuServe, as part of the package of services it offered its subscribers, maintained a general information service known as CompuServe Information Service ("CIS"). CIS included a variety of fora which provided an interactive platform for the discussion of over 150 topic areas. One of these topic areas was a journalism forum, which included a daily newsletter devoted to discussion of journalist activities and broadcast journalism, known as "Rumorville." Rumorville was published by Don Fitzpatrick Associates and was made available to CompuServe through a contract for service.[107] Additionally, the journalism forum was handled by a third party, Cameron Communications, Inc. Cameron, by contract with CompuServe, was required to, "'manage, review, create, delete, edit and otherwise control the contents' of the [j]ournalism [f]orum 'in accordance with editorial and technical standards and conventions of style as established by CompuServe.' "[108] CompuServe contractually disclaimed all rights to directly edit Rumorville.

{36} Cubby, Inc. also developed a computer database program called "Skuttlebut," which was dedicated to the discussion of news and gossip about broadcast journalists. Cubby asserted that Rumorville published defamatory comments about Skuttlebut. The comments stated that Skuttlebut's authors received their information in an unethical manner and described the forum as a " 'new startup scam.' "[109] Cubby filed suit against CompuServe and the individual author of Rumorville, essentially claiming a defamation cause of action.[110]

{37} The court, in response to a motion for summary judgment made by CompuServe, found that the service provider was only a distributor, who maintained no editorial control and thus dismissed CompuServe as a defendant.[111] The court explained that CompuServe, as a mere passive conduit, without direct editorial control, could not be held liable. The court found that, in spite of CompuServe's contractual relationship with Cameron and Rumorville, and in spite of the promulgation of guidelines by CompuServe for the proper management of its fora, CompuServe maintained no editorial control and had no knowledge or reason to have knowledge of the defamatory content.[112] The court analogized the information service to the maintenance of a for-profit library, thus establishing a high burden for the plaintiff.[113]

{38} In granting summary judgment, the effect of the court's decision was to establish a three-prong test. The court first determined if CompuServe was a publisher or distributor. This inquiry asked whether there was editorial control, or alternatively, if CompuServe was similar to a library or bookstore (i.e., a distributor of information).[114] The court, finding CompuServe to be a distributor, next determined if the defendant could be held liable as a distributor. To do so, the court required that once CompuServe had shown that it did not have actual knowledge of the defamatory material, the plaintiff must make an affirmative showing that the defendant either did know, or should have known, the defamatory nature of the statement. [115] Finally, the court inquired whether or not CompuServe could be held vicariously liable for the defamatory comments due to its contractual relationship with Cameron and Rumorville. The court rejected vicarious liability as a means of recovery, requiring that there be some showing of actual direction and control by CompuServe.[116]

2. Stratton Oakmont, Inc. v. Prodigy Services Co.[117]

{39} In a case strikingly similar to Cubby, the Supreme Court of New York held Prodigy, an Internet service provider, liable for statements made on an Internet bulletin board maintained for its customers.[118] In this case, a defamatory statement was posted to Prodigy's "money talk" bulletin board, claiming that Stratton Oakmont and its president, Daniel Porush, "committed criminal and fraudulent acts in connection with an initial public offering of stock," and other related claims of criminal activity. [119] Stratton Oakmont and Daniel Porush filed an action for libel along with nine other causes of action, and sought partial summary judgment against Prodigy.[120] The court was asked to determine, inter alia, "whether Prodigy may be considered a 'publisher' of the [defamatory] statements[,] and, whether Epstein, the [contracted] Board Leader for the computer bulletin board on which the statements were posted [by an unknown party], acted with actual and apparent authority as Prodigy's 'agent' for the purposes of the claims in this action."[121]

{40} The facts of the case revealed that Prodigy advertized itself as a "family oriented computer network," that "held itself out as a[n] online service that exercised editorial control," and "expressly liken[ed] itself to a newspaper."[122] Furthermore, Prodigy promulgated extensive guidelines describing what would be appropriate in postings and when a posting would be removed by its board leaders.[123] Prodigy also employed a screening program, which automatically screened content for "offensive material," and they employed board leaders to enforce its guidelines.[124] Though Prodigy claimed to have changed its policy on the automatic content review of postings, the court found no substantial evidence of such a change, and therefore, held Prodigy accountable to these prior statements of policy.[125]

{41} The court in Stratton also begins its analysis of the dispute by inquiring whether or not Prodigy may be considered a publisher or distributor of the defamatory material.[126] The Stratton court distinguishes Cubby and finds for the plaintiffs. [127] The court based its holding on two primary findings: first, that Prodigy held itself out to have content control, and second, that such control was actually performed through the use of an automatic software screening program and the "[g]uidelines which Board Leaders are required to enforce."[128] It determined that these practices constituted editorial control and thus, made Prodigy a publisher instead of a distributor. Further, the court found that the Board Leaders were agents of Prodigy because they were given direct guidelines which must be followed and given no authority to create policy.[129]

{42} The Stratton court, like in Cubby, again follows a three-step review. First, the court determines whether Prodigy is a publisher or a distributor. It explains that one who republishes a defamatory statement has the same liability as the original author.[130] The Stratton court finds editorial control through the power of Board Leaders to publish or remove postings. [131] The court next establishes a unique manner to determine why Prodigy is a publisher, by focusing on its public statements of editorial control, the use of automated computer software, and contracting for "editorial" decisions through its Board Leaders.[132] Finally, it also employs a form of vicarious agent liability rejected by the Cubby court.[133] The court, therefore, while claiming to be in harmony with the decision in Cubby, Inc. v. CompuServe, Inc.,[134] provides an apparently conflicting analysis to online service provider liability.

3. Communications Decency Act of 1996 § 230(c)[135]

{43} Shortly after the Stratton decision, Congress established a strong service provider protection from online defamation in the Communications Decency Act of 1996 ("CDA").[136] Subsection (c)(1) of this section provides, "[n]o provider or user of any interactive computer service shall be treated as the publisher or speaker of any information provided by another information content provider."[137] This mandate, in combination with subsection (2) of the same section,[138] provides a complete shield from a defamation suit for an online service provider, absent an affirmative showing that the service was the actual author of the defamatory content. Legislative intent shows that the section was passed as a specific response to the finding of liability in the Stratton case.[139]

C. Current Approach

1. Zeran v. America Online, Inc.[140]

{44} After the enactment of the CDA, the judicial approach to defamation cases changed significantly. In Zeran, the Fourth Circuit Court of Appeals gave significant force to the protections established in Section 230 of the CDA. The court found that this new statutory provision prohibited Zeran's action against America Online ("AOL"). [141]

{45} Kenneth Zeran brought an action against AOL for unreasonable delay in removing defamatory messages posted by a third party and for refusing to post a retraction.[142] The action was precipitated by a posting to an AOL bulletin board on April 25, 1995, by an unidentified party who advertized that the plaintiff was selling T-shirts making jest of the Oklahoma City Federal Building bombing.[143] The posting directed people to contact the plaintiff at his home phone number in Seattle, Washington to purchase the t-shirts. As a result of this posting and subsequent similar postings, Zeran received "a high volume of calls, comprised primarily of angry and derogatory messages, but also including death threats."[144] Zeran contacted AOL to have the messages removed, a retraction posted, and the account closed from which the postings were being made. The parties disputed how long the delay was in removing the defamatory postings, but the evidence is clear that AOL refused to post a retraction.[145] Subsequently, an Oklahoma City radio station broadcast the contents of the posting.[146] AOL filed a motion for judgment on the pleadings, which was granted in its favor and upheld upon appeal.[147]

{46} The Fourth Circuit rejected Zeran's argument that AOL should be held liable for failing, upon being given notice, to remove the defamatory material, to notify its customers that the statements were false, and to screen for future similarly false statements. The court stated that, "[b]y its plain language, § 230 [of the CDA] 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."[148] Thus, the court held that failure to exercise editorial control was not a basis for recovery after the passage of CDA § 230(c).[149] The court later rejected Zeran's claim that the provisions of CDA § 230(c) only protected publishers, but not distributors. The court explained that a distributor was just a sub-class of the legal classification "publisher," and therefore, would be covered under the liability protections of CDA § 230(c).[150] The court justified its broad interpretation of "publisher" by explaining, "'[t]he rule that statutes in derogation of the common law are to be strictly construed does not require such an adherence to the letter as would defeat an obvious legislative purpose or lessen the scope plainly intended to be given to the measure.'"[151] Finally, the court rejects liability based upon notice and limits Zeran's potential recovery to a suit against the original author of the posting.[152]

{47} The court makes four essential holdings in the Zeran case. First, the court notes that CDA § 230(c) bars Internet service provider liability for acting as either a publisher or distributor of defamatory statements.[153] The court then finds that this total protection for service providers was the explicit intent of Congress, who was seeking to overturn the rule established in Stratton.[154] Additionally, the court established, as a matter of law, that notice to an Internet service provider of defamatory content in the information it provides does not create liability.[155] Finally, the court recognizes that, in situations similar to Zeran's, the only party which may be held liable is the original author of the defamatory statement.[156]

2. Blumenthal v. Drudge[157]

{48} In this recent interpretation of the limits of protection afforded by CDA § 230(c), the United States District Court for the District of Columbia barred recovery for a plaintiff relating to defamatory statements, where the Internet service provider acted in an editorial role prior to posting of the statement.[158] The court recognized this vast protection while noting that under traditional defamation law, AOL, the service provider who "published" the defamatory statements, would have been held accountable.[159] Further, the court, as in Cubby and through the effect of CDA § 230, rejected the notion that contractually provided content would give rise to Internet service provider liability.[160]

{49} Matt Drudge was the author of an electronic publication known as the Drudge Report, which focused on gossip relating to the entertainment industry and governmental officials. Drudge disseminated his publication by e-mail, his personal World Wide Web site, and through AOL, with whom he had a contract to provide his information. [161] Drudge received a flat monthly fee of three thousand dollars from AOL for providing his publication to disseminate to AOL customers. [162] AOL did, however, reserve the right to modify the content of the report to meet AOL's standards. [163] The August 10, 1997 edition of the Drudge Report contained allegedly defamatory statements regarding Sidney and Jacqueline Blumenthal. The statements claimed that Sidney Blumenthal had a history of domestic violence against his wife. [164] Both Blumenthals were employees of the United States Government in Washington D.C., and were associated with the Office of the President. After publishing his report, Drudge immediately received a letter from the Blumenthals' attorney, which caused him to post a retraction on August 12, 1997.[165] The Blumenthals filed suit against both Drudge and AOL. The court held that, pursuant to CDA § 230, AOL was entitled to summary judgment.

{50} The Blumenthal court made several key findings and established a number of broad-ranging implications to CDA § 230, in protecting AOL from liability. The court found that, although Drudge was a contract employee of AOL, this fact did not provide liability unless a direct employee of AOL was the author of the defamatory statement.[166] Though the court acknowledged that, under traditional defamation law, AOL could be held liable as either a "publisher" or "distributor," even its retention of editorial control did not give rise to liability in the face of CDA § 230(c).[167] The court again makes the observation that the only potentially liable party is the original author, Matt Drudge.[168]

D. Suggested Approach: Original Author Liability, Substantive Editor - Service Provider Liability

{51} In defining the protections afforded by the CDA, the courts have nearly foreclosed the possibility of recovery for Cyber-Libel. Such an approach creates a special area of unique protection that can serve to encourage wild speculation without fear of liability. While Congress had laudable goals in attempting to encourage self-regulation of the Internet,[169] the protection provided has eliminated the possibility of substantial recovery, even where an Internet service provider retains editorial control over the defamatory content.[170] This Cyber-Libel scheme eviscerates the protections and disincentives to information sources in ensuring that false statements are not disseminated. The court is short-sighted in completely foreclosing recovery against Internet "publishers," as it ignores the dual interests of states in protecting the reputation of its domicilliaries and protecting its citizens from the dissemination of false information.[171]

{52} A more reasoned approach would find a more balanced treatment of the desired protection for Internet service providers and the traditional reputation protection of defamation law. It would provide substantive protection for online information distributors, while not shielding the same companies from gross displays of indifference, or even participation in the making of defamatory statements. The ideal approach would maintain the CDA §230(c)'s desired encouragement of self-regulation [172] without foreclosing recovery to plaintiffs who are grievously harmed by false statements.

{53} One logical approach would begin with the one constant between traditional defamation law and Cyber-Libel, as it has developed. This constant is the continued liability of the original author of the defamatory statement. [173] This would ensure in a large number of cases that at least some means of satisfaction, even if limited, is achieved. [174]

{54} An additional source of recovery should also be established in a limited number of circumstances. This recovery should be afforded where the Internet service provider has served as an active editor of the defamatory statement that it published. More precisely, the service provider should be held accountable where it takes the role of actively editing, either by adding additional text or by editing for length and content. Essentially, by this provision, content provided by a third party, but edited by the service provider, such as in an online magazine, is treated in the same fashion as traditional print media. The service provider is not to be subject to liability as a passive conduit, [175] or for automated software screening for objectionable material. [176] Additionally, where the control exercised by the service provider consisted of merely deciding whether to publish or to remove information, the provider is protected in accordance with CDA § 230(c). [177] Thus, the stated desire of Congress to overturn the rule in Stratton is preserved, while allowing for full recovery by a plaintiff, where the service provider actively participated in the defamation.

{55} This second source of recovery is not in conflict with the language of CDA § 230(c). Under § 230(c)(1), an Internet service provider who actively participates in the editing or writing of a defamatory statement is not merely a recipient of "information provided by another information content provider." [178] The service provider is, instead, deemed a co-author of the defamatory content, and therefore, would not be in the class of entities protected by the CDA. Furthermore, if the service provider engaged in active editing or substantive additions to the defamatory statement, it would no longer be acting in "good faith." [179] Recovery in such a scenario would, therefore, not be in direct violation with the statutory language passed by Congress.

{56} A more equitable median can be struck. Defamation law seeks to redress what is often a very painful, if not externally obvious, injury. The near preclusion of recovery under current interpretive maxims seems to overlook the very foundational precepts underlying defamation. It is as if neither the judicial nor legislative branch places value in a person's honor or reputation. This small adjustment in interpretation suggested above restores some measure of protection.


{57} Any truly comprehensive treatment of defamation law must address one last area of concern - retraction statutes. These statutes provide either a partial or complete defense to the defamation action, depending upon the jurisdiction involved. [180] As a means of protecting the publisher in a print world, requiring a fair rebuttal and retraction of the defamatory content previously published can be a means of providing - in some manner - a restoration of the injured party's reputation. In this sense, the retraction not only serves the interest of truth, but defends the previously tattered honor of an individual. This makes the retraction potentially the most effective means of remedying injury at the earliest possible moment. One need look no further than the local newspaper to see, almost daily, the exercise of this very principle. [181]

{58} Though the retraction in a traditional print media can be very effective, it is not necessarily as effective in the online world. While many web-surfers do reach the same page daily, [182] many recreational Internet users may cross a page only once. Thus, the applicability of the retraction may be legitimately questioned. However, when a particularly offensive statement or communication is carried on a webpage, a strong retraction can often prove to quell the damage done. [183] Thus, the web-surfer who is more than just a casual passerby to the particular website will be informed. It is with this notion of at least partial effectiveness, and the fact that at a minimum, thirty-three states have retraction statutes, [184] that the issue must be addressed.

A. Common Law

{59} At common law, retraction was a factual basis by which the publisher could mitigate damages resulting from a defamatory statement.[185] Mere retraction was generally insufficient to avoid liability on the part of the publisher, but was seen as a form of good faith attempt by the publisher to lessen the effect of the libelous material. Retraction under common law was also given the vague limitation of requiring that it must be sufficient in light of what a reasonable person would understand as satisfactory, given the surrounding circumstances.[186] The level of mitigation that the retraction caused would be determined by the fact-finder upon deciding that sufficiency had been achieved.[187] The lack of a retraction could be viewed as evidence of malice in determining whether or not punitive damages could be awarded. Thus, while a publisher is under no affirmative duty to publish a retraction, such retraction can be used to avoid punitive damages and to ease the general damage burden.

B. Current Approach/Statutory

{60} As previously noted, the majority of jurisdictions have codified some form of the common law retraction defense. Seemingly endless machinations have been employed to create mildly varying schemes in relation to retraction. These sometimes conflicting approaches often "muddy the waters" of establishing a clear defense, particularly when one is a publisher or broadcaster with a wide distribution or reception. If the "publisher" of the defamatory statement is even as common as a local television station, application of retraction statutory mechanisms can be difficult. One such problematic example can be extrapolated from a hypothetical event in our nation's capital. If, for example, a Washington D.C. network affiliate broadcasts false rumors about a local entertainer of little note, it is possible that the station, in seeking to employ retraction principles, may be faced with as many as five different rules. [188] In light of this potential confusion relating to traditional media sources, it is easy to imagine the near impossibility of predicting the necessary requirements for retraction in a Cyber-Libel case. A beginning point involves the need to identify the primary areas of differentiation.

{61} Differentiation can arise in no less than five major areas. First, divergence can be found in the damages recoverable after a successful retraction. Damages may be limited to actual damages, [189] or mitigated by the retraction, [190] while others may only preclude punitive damages upon a retraction. [191] Variation can be found in whether the jurisdiction requires a retraction demand on the part of the plaintiff and the effect of this demand. [192] A good number of states require that the retraction be made within a specific time period, [193] while others provide exacting requirements of form and location for the retraction. [194]

{62} This mishmash of variety only serves to guarantee confusion for a potentially liable party. A more uniform and flexible standard needs to be developed in order to provide a modicum of clarity and to protect the Internet service provider or publisher. While many are loathe - and with good reason - to encourage further federal legislation in the Cyber-Libel area of law, retraction may be the one area where legislation could be beneficial.

{63} First and foremost, any legislation must establish a reasonable but adaptable standard for determining when a sufficient retraction has occurred. This principle in seeking fairness can, and should, use a tool as simple as human wisdom. One possible touchstone is that standard established by California of "substantially conspicuous." [195] Put simply, if the retraction is done in a manner that is open, noticeable, and not of unnecessary variance in visual importance on the particular website, it should be considered sufficient. [196] In the interest of providing real protection for the "publisher," while not eliminating recovery for the potential plaintiff, it is suggested that the retraction have the effect of partially limiting damages. This balance between maintaining some right to recovery, and affording protection for the publisher who has simply made an honest mistake, can be achieved while retaining a level of discretion. A party who makes a sufficient retraction should not be liable for punitive damages, as they have demonstrated by their retraction that it is unlikely conventional malice [197] actually exists. Further, as has been previously explained, such retraction could serve to lessen the impact of the defamatory statement. Accordingly, retraction should be accorded status as a mitigating factor when determining damages. Lastly, a formal demand for retraction should not be required. The establishment of a formal demand seems without reasonable foundation. Some notice should be required; however, it seems to be not unduly burdensome to the plaintiff to require that, at a minimum, he give notice of the falsity of the defamatory statement to the online publisher.

{64} This formula provides a simple means of applying retraction principles to a modern online world. There is predictability and convenience, while protecting both parties' interests. The defamed party not only has the potential impact lessened by receiving the benefits of encouraging retraction, but he is also not forestalled from all recovery simply because of a footnote at the bottom of a webpage. The Internet publisher receives protection through the guaranteed avoidance of punitive damages, and is rewarded with mitigation if prompt and "substantially conspicuous" action is taken. While it may still be left to the jury to decide the end effect of the retraction's mitigation power, on the whole, it can be expected that swift and appropriate retraction should significantly reduce liability exposure.


{65} Though much has been written about the development of Cyber-Libel law, no unified rational approach has been developed by the courts to handle the variety of complex issues that arise in this unique area of law. By providing a direct course for courts to follow in an area of the law which is likely to be confusing for the average jurist, the Supreme Court provides the greatest of services. The author has humbly presented a course of action that attacks the primary areas of concern in a Cyber-Libel case - from online jurisdiction - through a suggested modification of the interpretation of the CDA § 230 - to the applicability of retraction statutes. The approach takes into consideration the concerns of both plaintiff and defendant, while remaining mindful of congressional directives. The author suggests that jurisdictional questions in Cyber-Libel cases be controlled by the Zippo [198] nature of the communications test, an already prevalent approach that provides the greatest fairness and accuracy. Choice of law in Cyber-Libel is best served by the simplicity of domicilliary focus, as found in the Restatement section 150, [199] augmented by the availability of a modified effects test, where a party pleads that domicile is not controlling. [200]

{66} It is suggested that substantive Cyber-Libel law be brought to a reasoned middle ground between the current bar on service provider recovery and the interest in plaintiff recovery. Any reaffirmation of the importance of freedom from defamatory attack must include some movement away from the current, rigid, near bar on recovery. This balancing is best achieved by employing a system of liability that opens service providers liability for defamatory statements, when the provider has substantively edited the statement, while not opening liability for minor editing, existence as a mere conduit, or electronic-filtering activities. [201] Furthermore, the application of retraction statute-like defenses on behalf of the service provider should be encouraged as a means of early mitigation and reputation protection.

{67} There can be no injury more personal than defamation of character. The pride and self worth of an individual often go beyond the bounds of mere monetary loss. It is hoped that the scheme provided herein would serve to protect the sensitive nature of a defamation plaintiff, while still providing significant protection to information sources. The balance is a difficult one to strike, but some "happy mediun" must be found.


Barry J. Waldman

[*] Barry Waldman received a B.A. History from the University of Virginia in 1995 and received a J.D. from the University of Richmond, School of Law in 1999. Mr. Waldman is a Litigation Associate with the law firm of Trichilo, Bancroft, McGavin, Horvath & Judkins, P.C. in Fairfax, Virginia.


[**]. NOTE: All endnote citations in this article follow the conventions appropriate to the edition of THE BLUEBOOK: A UNIFORM SYSTEM OF CITATION that was in effect at the time of publication. When citing to this article, please use the format required by the Seventeenth Edition of THE BLUEBOOK, provided below for your convenience.

Barry J. Waldman, Comment, A Unified Approach to Cyber-Libel: Defamation on the Internet, a Suggested Approach, 6 RICH. J.L. & TECH. 9 (Fall 1999), at


[1]. Zippo Mfg. Co. v. Zippo Dot Com, Inc., 952 F. Supp. 1119, 1123 (W.D. Pa. 1997) (quoting Panavision Int'l L.P. v. Toeppen, 938 F. Supp. 616 (C.D. Cal. 1996)).

[2]. See American Civil Liberties Union v. Reno, 929 F. Supp. 824, 830-848 (E.D. Pa. 1996).

[3]. The "World Wide Web" (also known commonly as the "Web") is "[a] hypertext-based system for finding and accessing Internet resources," which contains various text and graphic based electronic pages containing information on subjects selected by the "webmaster" (one who authors and/or maintains the "web-page"). Paul Jacobsen, Net Law: How Lawyers Use the Internet 223 (1997).

[4]. Zippo, 952 F. Supp. at 1123-24.

[5]. See, e.g., Robert M. O'Neil, The Drudge Case: A Look at Issues in Cyberspace Defamation, 73 Wash. L. Rev. 623 (1998); Bruce W. Sanford & Michael J. Lorenger, Teaching an Old Dog New Tricks: The First Amendment in an Online World, 28 Conn. L. Rev. 1137 (1996); David R. Sheridan, Zeran v. AOL and the Effect of Section 230 of the Communications Decency Act Upon Liability for Defamation on the Internet, 61 Alb. L. Rev. 147 (1997); R. James George, Jr. & James A. Hemphill, Defamation Liability and the Internet, in 18th Annual Institute on Computer Law, at 691 (PLI Patents, Copyrights, Trademarks, and Literary Property Course Handbook Series No. 507, 1998); Douglas B. Luftman, Note, Defamation Liability for On-Line Services: The Sky Is Not Falling, 65 Geo. Wash. L. Rev. 1071 (1997); Sarah Beckett Boehm, Note, A Brave New World Of Free Speech: Should Interactive Computer Service Providers Be Held Liable for the Material They Disseminate?, 5 Rich. J. L. & Tech. 2, (Winter 1998) available at (visited Sept. 24, 1999) <>.

[6]. For the purposes of this paper Cyber-Libel, Online Defamation, and Internet Libel will be used synonymously to indicated defamatory statements made over the Internet or through other Computer-related technological means.

[7]. For the purposes of this article, the author assumes that the case is to be litigated in a federal court sitting subject to its diversity jurisdiction pursuant to 28 U.S.C. §1331 (1994).

[8]. See World-Wide Volkswagen Corp. v. Woodson, 444 U.S. 286, 291 (1980).

[9]. Patriot Systems, Inc. v. C-Cubed Corp., 21 F. Supp.2d 1318,1320 (D. Utah 1998) (citation omitted).

[10]. EDIAS Software Int'l, L.L.C. v. Basis Int'l, Ltd., 947 F. Supp. 413, 417 (D. Ariz. 1996) (citation omitted).

[11]. See id. at 416.

[12]. See, e.g., Cybersell, Inc. v. Cybersell, Inc., 130 F.3d 414, 416 (9th Cir. 1997).

[13]. Keeton v. Hustler Magazine, Inc., 465 U.S. 770, 774 (1984) citing World-Wide Volkswagen, 444 U.S. at 297-98, and International Shoe Corp. v. Washington, 326 U.S. 310, 317 (1945).

[14]. EDIAS Software Int'l, 947 F.Supp. at 416 (comparing Helicopteros Nacionales v. Hall, 466 U.S. 408 (1984) (explaining test for general jurisdiction), and Burger King v. Rudzewicz, 471 U.S. 462 (1985) (explaining test for specific jurisdiction)).

[15]. Helicopteros, 466 U.S. at 414-16.

[16]. See id. (denying general jurisdiction over a corporate defendant who had purchased millions of dollars of goods, negotiated contracts and trained employees in the forum state). General jurisdiction is most often employed without requiring a full analysis when the defendant individual/entity is "domiciled" in the forum state.

[17]. See Burger King, 471 U.S. at 472-73.

[18]. EDIAS, 947 F. Supp. at 417 (quoting Ballard v. Savage, 65 F.3d 1495, 1498 (9th Cir. 1995)).

[19]. Patriot Systems, 21 F. Supp.2d at 1321 (quoting Far West Capital, Inc. v. Towne, 46 F.3d 1071, 1074 (10th Cir. 1995)).

[20]. See World-Wide Volkswagen, 444 U.S. at 297.

[21]. See, e.g., Zeran v. American Online, Inc., 129 F.3d 327 (4th Cir. 1997) (finding that an anonymous third party posted a defamatory statement on AOL).

[22]. A "hacker" is an individual who uses computer and/or Internet technology to gain access to, and often destroy protected information of another.

[23]. See Jonathan D. Hart, et al., Cyberspace Liability, in Libel & Newsgathering Litigation-Getting & Reporting The News 1998, at 139 (PLI Patents, Copyrights, Trademarks, and Literary Property Course Handbook Series No. 523, 1998).

[24]. See id.

[25]. 947 F. Supp. 413 (D. Ariz. 1996).

[26]. See id. at 421.

[27]. See id. Though the court focuses primarily on an "effects test" (see infra Section II (B)(2)), the court employed a totality approach to determine sufficiently purposeful availment.

[28]. No. CV 97-4070-WDK, 1997 WL 835442 (C.D. Cal. 1997) (not reported in F. Supp.).

[29]. See id. at *3.

[30]. Id.

[31]. Id.

[32]. 130 F.3d 414 (9th Cir. 1997).

[33]. See id. at 419-20.

[34]. Hart, et al., supra note 24, at 145 (quoting Calder v. Jones, 465 U.S. 783, 788 (1984)).

[35]. See EDIAS, 947 F. Supp. at 420 (D. Ariz. 1996). The author notes that the jurisdictional analysis in EDIAS can be viewed as either an effects test approach, a totality of contacts approach, or a combination thereof.

[36]. See id. at 419-20.

[37]. Id. at 420.

[38]. 631 F. Supp. 1356, 1363 (C.D. Cal. 1986).

[39]. See EDIAS, 947 F. Supp. at 420.

[40]. See generally Keeton v. Hustler Magazine, Inc., 465 U.S. 770, 773-74 (1984) (finding that single publication rule did not bar plaintiff's suit for damages incurred by nationwide publication of defamatory material). The single publication rule allows only one action for damages for any single publication. The single publication rule is an exception to the "'general rule that each communication of the same defamatory matter by the same defamer, . . . is a separate and distinct publication, for which a separate cause of action arises.'" Id. at 774 n.3 (citation omitted).

[41]. Id. at 773-74.

[42]. Id. at 774, 777 (citing Restatement (Second) of Torts § 577A cmt. a (1997)).

[43]. See id. at 777.

[44]. 977 F. Supp. 404 (E.D. Va. 1997).

[45]. See id. at 408.

[46]. Id. Furthermore, the court noted that one news forum where the press release was carried specifically stated that the information would be received in Virginia. See id.

[47]. For further background on online jurisdiction in light of the TELCO decision, see generally Donnie L. Kidd, Jr., Note, Casting the Net: Another Confusing Analysis of Personal Jurisdiction and Internet Contacts in TELCO Communications v. An Apple A Day, 32 U. Rich. L. Rev. 505 (1998) (offering helpful background information about jurisdictional concerns germane to the Internet).

[48]. See A. Michael Froomkin, Anonymity and Its Enmities, 1995 J. Online L. art. 4 (visited Sept. 21, 1999) <> (discussing Internet anonymity). While it is true that many websites and e-mail servers keep a record of the communications made between parties and the identity of parties, these records are often kept in a less substantial manner than paper communications and are often subject to periodic database purges. Additionally, with the still persistent problems in identifying individual "web surfers" who wish to remain relatively anonymous, determining if a particular individual in a particular forum has accessed information contained on the Internet can be tedious at best, and impossible in some instances. The author, therefore, rejects the requirement of actual numerical identification of the electronic contacts made by a defendant with persons or entities within the forum and instead suggests that the nature of the online communication should be determinative.

[49]. See Keeton, 465 U.S. at 773-74.

[50]. 952 F. Supp. 1119 (W.D. Pa. 1997).

[51]. Id. at 1124 (citations omitted).

[52]. See id. Where the libelant posts the defamatory statement to a passive website the author wishes to express the opinion that the suit may still be brought in the jurisdiction from which the statement was composed. Upon composition, the statement is "delivered" to the computer network which will then "post" it to the Internet. Under this same analysis, a plaintiff might argue that if the computer network which acts as the "server" for the libelant's website is located in a different forum than the where the libelant composed the message, jurisdiction may also be proper in the forum where the "server" is located. The author limits the focus of this analysis to Internet activities and expresses no opinion on Intranet defamation.

[53]. But see TELCO, 977 F. Supp. at 406-08 (explaining that a passive website alone, which serves as an advertisement, may establish jurisdiction for the purposes of defamation).

[54]. See supra notes 35-37 and accompanying text.

[55]. For example, if the website were one maintained by a newspaper, a television station or the like, the website would fall under that category of "clearly doing business" in any state which it reached. The website would exist as part of the libelant's core business and be considered a secondary means of delivering its product.

[56]. See Zippo, 952 F. Supp. at 1124. See also Cybersell, 130 F.3d at 419; Patriot Sys., 21 F. Supp.2d at 1323.

[57]. TELCO, 977 F.Supp. at 407.

[58]. See supra note 19 and accompanying text.

[59]. For the purposes of this article, choice of law and conflicts of law will be used synonymously to indicate the judicial decision of which state law to apply in a suit brought subject to the diversity jurisdiction of the federal courts.

[60]. See Corey B. Ackerman, World-Wide Volkswagen, Meet the World Wide Web: An Examination of Personal Jurisdiction Applied to a New World, 71 St. John's L. Rev. 403 (1997); Jason L. Brodsky, Civil Procedure - Surfin' The Stream Of Commerce: Compuserve v. Patterson, 89 F.3d 1257 (6th Cir. 1996), 70 Temple L. Rev. 825 (1997); and Kidd, supra note 47.

[61]. See Hart, supra note 24, at 158 (explaining that a unified approach to choice of law questions in Cyber-Libel has not been addressed by the courts, and expressing the concern that an adequate regime must be established to avoid a chilling effect on speech).

[62]. See, e.g., Va. Code Ann. § 8.01-45 (Michie 1994); Cal. Civ. Code § 44 (West 1996); N.Y. Civ. Rights Law § 74 (McKinney 1997).

[63]. See supra Section V (offering more information on retraction statutes).

[64]. In general, the first inquiry in any conflicts of law issue is to determine whether the various state laws to be employed have any substantial conflict. If there is no substantial conflict between the forum's substantive law and the other potential sources of substantive law, the court will typically employ the substantive law of the forum without further choice of law analysis. See Wilson v. Slatalla, 970 F.Supp. 405, 413 (E.D. Pa. 1997).

[65]. Restatement (Second) of Conflicts of Laws § 145 (1971). The author notes that not every jurisdiction has adopted the Restatement approach to conflicts of laws analysis. The Restatement approach, however, is commonly used. For the purposes of this article, the forum will be assumed to have adopted the Restatement.

[66]. Id. § 6.

[67]. MacDonald v. General Motors Corp., 110 F.3d 337, 342 (6th Cir. 1997).

[68]. See Gann v. Fruehauf Corp., 52 F.3d 1320, 1325 (5th Cir. 1995).

[69]. MacDonald, 110 F.3d at 342.

[70]. See, e.g., Erie R.R. Co. v. Tompkins, 304 U.S. 64 (1938) (finding state common law will control in substantive matters, but federal rules and procedures will govern matters that are procedural in nature, and that the conflict of laws rules of the state where the federal court is located must be followed). This conflicts analysis is unnecessary in determining what procedural law is to be applied, as the procedural law of the forum is used in nearly all cases. The court will, therefore, employ the choice of law scheme of the forum state, as this portion of the law is generally considered procedural. See Klaxon Co. v. Stentor Elec. Mfg. Co., 313 U.S. 487, 496 (1941).

[71]. See Restatement (Second) Of Conflicts of Laws § 150 (1971).

[72]. Id. § 150(1). Arguably, online defamation could fit within the definition of "aggregate communication." The author, however, argues that such an application may not be totally appropriate as the comments and reporter's notes to Section 150 do not reference online communication. The author argues for a modified approach in Section III(B) of this survey.

[73]. Id.

[74]. Id. § 145.

[75]. See supra notes 65 & 66 and accompanying text.

[76]. Restatement (Second) of Conflicts of Law § 150(2).

[77]. See id. § 150(3).

[78]. See Wilson, 970 F.Supp. at 414; Fornshill v. Ruddy, 891 F. Supp. 1062, 1069 (D. Md. 1995).

[79]. See Wilson, 970 F. Supp. at 415 (citing Lamelza v. Bally's Park Place, Inc., 580 F. Supp. 445, 446-47 (E.D. Pa. 1984)).

[80]. See supra note 48 and accompanying text.

[81]. See Restatement (Second) of Conflicts of Laws § 150 cmt. b (1971) ("[T]his [s]ection calls for application of the local law of the state selected pursuant to the provisions of [s]ubsection (2) and (3), [the determination of domicile of the plaintiff].").

[82]. Wilson, 970 F. Supp. at 414 (quoting Fitzpatrick v. Milky Way Prods., Inc., 537 F. Supp. 165, 171 (E.D. Pa. 1982)).

[83]. But see Rome v. Glanton, 958 F. Supp. 1026, 1034 n.5 (E.D. Pa. 1997) (declaring that the state of domicile is the state of most significant interest).

[84]. For example, a small to mid-sized computer manufacturing company may be registered in Delaware to take advantage of that state's pro-business rules of incorporation, yet maintain its production facility in Texas and distribute its products in the southwestern United States, and not make a single sale in Delaware. Thus, any defamation of its product would have no effect in Delaware as it would not lead to decreased sales in that forum.

[85]. For example, a defamatory statement made in Washington, D.C. against a mid-level governmental appointee may never reach the appointee's home state, but could cause reputation damage in his place of employment and his temporary location of residence, Washington, D.C.

[86]. 947 F. Supp. 413 (D. Ariz. 1996).

[87]. See supra notes 35, 36 & 37 and accompanying text.

[88]. See supra notes 37 & 38 and accompanying text (purposefully directing e-mail to a particular forum would provide a strong argument that the particular forum has the most significant interest in protecting its citizens from false statements and the defamed plaintiff from injury).

[89]. See Rodney A. Smolla, Dun and Bradstreet, Hepps and Liberty Lobby: A New Analytic Primer on the Future Course of Defamation, 75 Geo. L.J. 1519 (1987) (offering a more complete discussion of traditional defamation law and its development).

[90]. The author and many other commentators mark the beginning of modern constitutional defamation law with New York Times Co. v. Sullivan, 376 U.S. 254, 280 (1964).

[91]. Black's Law Dictionary 417 (6th ed. 1990).

[92]. See Restatement (Second) of Torts § 558(a) (1977).

[93]. See id. § 559.

[94]. See id. § 558(b).

[95]. See id.

[96]. See id. § 558(d).

[97]. One who is so widely known that significant portions of their life are open to public examination. For example, a famous movie star, a top-level professional athlete or nationally known author. A limited public figure, one who injects himself into the public debate about a particular issue and is defamed in relation to this issue, will often receive the same burden as a pervasive public figure.

[98]. See Gertz v. Robert Welch, 418 U.S. 323 (1974).

[99]. See New York Times Co., 376 U.S. 254, 279-80, -85-86 (1964).

[100]. See Gertz, 418 U.S. at 340-41, -45-46.

[101]. See Restatement (Second) of Torts § 580B (1977).

[102]. See Douglas B. Luftman, Note, Defamation Liability For On-Line Services: The Sky Is Not Falling, 65 Geo. Wash. L. Rev. 1071, 1083 (1997).

[103]. See id.

[104]. See id. at 1084 (citing Restatement (Second) of Torts § 588 (1977)).

[105]. See id. at 1085 (citing Restatement (Second) of Torts § 581(1) (1977)).

[106]. Cubby, Inc. v. CompuServe, Inc., 776 F.Supp. 135 (S.D.N.Y. 1991).

[107]. See id. at 137.

[108]. Id. (citation omitted).

[109]. Id. at 138 (citation omitted).

[110]. See Cubby, 776 F.Supp. at 137, 138. Cubby filed an "action for libel, business disparagement, and unfair competition," resulting from the "allegedly defamatory statements contained in Rumorville." Id.

[111]. See id. at 141.

[112]. See id.

[113]. See id. at 140.

[114]. See generally id. at 138-141 (noting that publishers usually are liable for publishing defamatory material, but that libraries, book stores, and news vendors generally lack knowledge of the defamatory statements and, therefore, may not be liable).

[115]. See id. at 141.

[116]. See id. at 142-43.

[117]. 63 U.S.L.W. 2765 (N.Y. Sup. 1995). 1995 WL 323710 (N.Y. Sup.).

[118]. See Stratton Oakmont, Inc., 63 U.S.L.W. 2765 (pinpoint citations for this case throughout this paper will be made in accordance with the page numbering established by the reporting of the case in 1995 WL 323710).

[119]. Id. at *1.

[120]. See id.

[121]. Id.

[122]. Id. at *2.

[123]. See id.

[124]. See id.

[125]. See id. at *3.

[126]. See id.

[127]. See id. at *4-5.

[128]. Id. at *4.

[129]. See id. at *6-7.

[130]. Id. at *3 (citing Cianci v. New Times Pub. Co., 639 F.2d 54, 61 (2d Cir. 1980)).

[131]. See id. at *4-5.

[132]. See id. at *2-5.

[133]. See id. at *6-7.

[134]. 776 F. Supp. 135 (S.D.N.Y. 1991).

[135]. 47 U.S.C. § 230 (1998 Supp.).

[136]. See id.

[137]. Id. § 230(c)(1).

[138]. See id. § 230(c)(2)(A)-(B). Subsection (2)(A) and (B) protect the service provider from a finding of being a publisher through any action taken to censor or restrict access to "objectionable" content "whether or not such material is constitutionally protected." Id.

[139]. See Hart, supra note 24, at 134-35.

[140]. 129 F.3d 327 (4th Cir. 1997).

[141]. See id. at 328.

[142]. See id.

[143]. See id. at 329. On April 19, 1995, the Alfred P. Murrah Federal Building in Oklahoma City, Oklahoma was destroyed in a car bombing, killing168 individuals and injuring many others. See Steven K. Paulson, Nichols Lawyer Seeks New Trial, Wash. Post, July 8, 1999, at A160; see also 1999 WL 17013068.

[144]. Zeran, 129 F.3d at 329.

[145]. See id.

[146]. See id. The court's decision does not include an analysis of Zeran's action against the radio station as it later made an on-air apology. See id.

[147]. See Zeran v. America Online, Inc., 958 F. Supp.1124 (E.D. Va. 1997), aff'd 129 F.3d at 329, 335 (4th Cir. 1997).

[148]. Zeran, 129 F.3d at 330 (emphasis added).

[149]. See id.

[150]. See id. at 332-34.

[151]. Id. at 334 (citation omitted).

[152]. See id. at 330, 333. By limiting Zeran's recovery to the original author of the posting, the court effectively denied him recovery as the posting was anonymous, and AOL was unable to provide records to show the identity of the defaming party. See id. at 329 n.1.

[153]. See id. at 330.

[154]. See id. at 330-31.

[155]. See id. at 333. This holding seems to contrast with some of the principles of retraction law. See supra Section V for more detailed discussion of retraction law.

[156]. See Zeran at 330. Thus, because of the ability of Cyber-Libel tortfeasors to hide their true identity, injured parties like Zeran may go uncompensated.

[157]. 992 F.Supp. 44 (D.D.C. 1998).

[158]. See Blumenthal, 992 F. Supp. 44.

[159]. See id. at 51.

[160]. See id. at 51-52.

[161]. See id. at 47. The Drudge Report can also be found at Matt Drudge, Drudge Report (Visited Sept. 21, 1999) <>. It is interesting to note that the Drudge Report was the first media source to carry the news of President Clinton's now admitted affair with Monica Lewinsky.

[162]. See id.

[163]. See id.

[164]. See id. at 47-48. There appears to have been no real substantiation for this rumor.

[165]. See id.

[166]. See generally id. at 50-51 (noting that AOL would not be immune from suit under § 230(c)(1) if AOL was the creator of the information).

[167]. See id. at 50-53.

[168]. See id. at 51. The court did, however, reject Drudge's motion to dismiss for lack of personal jurisdiction, opening up the possibility of at least a limited recovery against Drudge as an individual. See id. at 58.

[169]. See Zeran, 129 F.3d at 331.

[170]. See generally Blumenthal, 992 F.Supp. 44 (holding that AOL may not be liable for defamatory content it provided and only allowing potential recovery against Drudge, the author, who has a stated annual income of $36,000 a year).

[171]. See, e.g., EDIAS, 947 F.Supp. at 420 (finding that defendant "should not be permitted to take advantage of modern technology through an Internet [w]eb[]page and forum and simultaneously escape traditional notions of jurisdiction").

[172]. See supra note 167 and accompanying text.

[173]. See Zeran, 129 F.3d at 330; see also, Black's Law Dictionary 417 (6th ed. 1990) (defining defamation as the "intentional false communication, either published or publically spoken, that injures another's reputation or good name").

[174]. See, e.g., Blumenthal, supra notes 166-68 (explaining that the Blumenthals could recover against the original author, Drudge, but would probably achieve only a limited actual recovery as Drudge, on a salary of $36,000 a year would not have the same substantial resources as America Online).

[175]. See supra note 102 and accompanying text (a passive conduit may not traditionally be held liable for defamatory material which passes through its service system).

[176]. But see Stratton, supra note 106.

[177]. But see id. at *3-4.

[178]. Communications Decency Act of 1996 § 509(c)(1), 47 U.S.C. § 230(c)(1) (1999 Supp.).

[179]. See id. § 230(c)(2)(A) (stating, "any action voluntarily taken in good faith to restrict access to or availability of material that [is] consider[ed] to be obscene, lewd, . . . or otherwise objectionable . . ." may not be the basis of liability) (emphasis added).

[180]. See generally John Francis Major, Sufficiency of Retraction of Defamatory Statement, 40 Am. Jur. Proof Of Facts 2d 649 (1984). (explaining generally that a retraction was not a defense to defamation. Rather, retraction is considered to mitigate damages). See id. at 666.

[181]. The author admits, however, that often these retraction deal with less intrusive subjects than the typical defamation case, and often the retraction is relegated to a less eye-catching location in the particular publication.

[182]. Most web browsers begin each session on a particular homepage, and often this page is selected by the particular user. Thus, it is this webpage that is guaranteed to get at least momentary time on a person's computer each session.

[183]. See, e.g., If Not the Devil, Did Atheism Make Her Do It?: "Dr. Laura" Fights Nudie Pics on the Net ... (last modified Nov. 8, 1998) <>; Dr. Laura Nekkid on the Net, Mr. Showbiz News (last modified Oct. 23, 1998) <>; Patrizia DiLucchio, Dr. Laura, How Could You? The Culture of Technology The Technology of Culture: Table Talk (last modified Nov. 3, 1998) <>. For example, when indiscrete pictures of Dr. Laura Schlesinger reached the Internet, they received extensive media coverage. However, quick action on the part of Laura Schlesinger reduced exposure, as the pictures were removed and a statement was posted within forty-eight hours. See id.

[184]. See Bruce W. Sanford, Libel and Privacy, App. B (2d. ed. 1991).

[185]. See, e.g., Donna M. Murasky, Avoidable Consequences in Defamation: The Common- Law Duty to Request a Retraction, 40 Rutgers L. Rev. 167 (1987).

[186]. See, e.g., Turner v. Hearst, 47 P. 129, 131-32 (Cal. 1896) (discussing how a retraction may mitigate damages depending on the sincerity and promptness of the apology).

[187]. See id.

[188]. Washington D.C. television and radio stations can be received with proper equipment and/or cable access in parts of Virginia, Maryland, the District of Columbia, Delaware, and West Virginia.

[189]. See, e.g., Mass. Ann. Laws ch. 231 § 93 (Law. Co-op. 1985).

[190]. See Sanford, supra note 184, at 593.

[191]. See, e.g., Conn. Gen. Stat. § 52-237 (1995).

[192]. See Sanford, supra note 184, at 594-95.

[193]. See, e.g., Neb. Rev. Stat. § 25-840.01 (1998) (prescribing that a retraction is effective only if published within three weeks of the defamatory publication); Ga. Code Ann. § 51-5-11 (1991) (prescribing that a retraction is effective only if published within seven days of the defamatory statement).

[194]. See, e.g., Okla. Stat. Ann. tit. 12, § 1446a-b (West 1980) (requiring an eighteen-point typeface, a heading indicating that it was a retraction, and that it be published "on the same page").

[195]. Cal. Civ. Code § 48a (West 1996).

[196]. Some weighing of the timeliness of the retraction may also prove important. In determining if the retraction is conspicuous, a temporal factor would be important. If the original defamatory statement is no longer recalled, then a retraction could be of little use in restoring honor or correcting a false statement.

[197]. Here, the author is referring to the common English definition of malice meaning enmity or dislike, and not its legal definition which relates directly to the constitutional tests established in First Amendment constitutional law.

[198]. See Zippo, 952 F. Supp. 1119 (W.D. Pa. 1997). See also supra Section II(C).

[199]. See Restatement (Second) on Conflicts of Law § 150 (1971).

[200]. See supra Section III(B).

[201]. See supra Section IV(D).

Copyright 1999 Richmond Journal of Law & Technology