Cite as "Neil Fried, Dodging the Communications Decency Act when Analyzing Libel Liability of On-line Services: Lunney v. Prodigy Treats Service Provider like Common Carrier, 1 COLUM. SCI. & TECH. L. REV. 1, (Nov. 29, 1999) ." Published on November 29, 1999. Dodging the Communications Decency Act when Analyzing Libel Liability of On-line Services: Lunney v. Prodigy Treats Service Provider like Common Carrier Rather than Address Retroactivity Issue by Neil Fried[*] I. Introduction P1. Despite provisions of the Federal Communications Decency Act that prohibit treating interactive computer services[1] as publishers of third-party speech, the Appellate Division of the N.Y. Supreme Court[2] applied a libel analysis in its December 1998 decision in Lunney v. Prodigy Servs. Co.[3] In December 1994, teenager Alex Lunney sued Prodigy for libel based on offensive e-mail and electronic bulletin board messages sent in September 1994 from Prodigy accounts fraudulently opened under his name.[4] The Appellate Division granted Prodigy summary judgment, but not before examining the elements of libel, drawing flawed analogies between service providers and telephone and telegraph companies, and extending to service providers a qualified, state common-law privilege historically applied to common carriers. The court appears to have done so to avoid reaching issues regarding retroactive application of the CDA. P2. This article discusses the Appellate Division’s analysis, and concludes that the court’s decision to sidestep the CDA and to compare service providers to telephone and telegraph companies was unwise. Part II.A summarizes traditional libel theory. Part II.B begins by discussing Cubby, Inc. v. CompuServe, Inc., one of the first significant judicial decisions to apply traditional libel analysis in the on-line context, and Stratton Oakmont, Inc. v. Prodigy Servs. Co., the case that prompted Congress to enact the libel-related provisions of the CDA. Part II.B then discusses the libel-related provisions of the CDA and two cases that have interpreted them: Zeran v. America Online Inc. and Blumenthal v. Drudge. Parts III and IV summarize and then analyze Lunney v. Prodigy. Part V makes recommendations to the N.Y. Court of Appeals,[5] which heard oral arguments October 13, 1999, in Lunney’s appeal. II. Libel A. An Overview P3. Libel is a tort of defamation, and is usually described as an injury to reputation caused by written communication.[6] The elements of libel generally include: (1) publishing, (2) to a third party, (3) a statement of and concerning the plaintiff, (4) that is defamatory, (5) false, and (6) unprivileged, (7) with some degree of fault by the defendant, (8) that causes the plaintiff injury.[7] 1. Publishing P4. The publication element is not to be taken literally, but simply requires some communication of the libelous material.[8] To have “published” a statement, however, a defendant must have intended—or reasonably been able to anticipate—the publication; inadvertent publication is not actionable absent negligence.[9] Because publication is often a collaborative process, liability for a libelous statement is not limited to its author.[10] Those who are in the business of making their facilities available to disseminate the [statements] may also … be regarded as publishers. They are intentionally making the contents available to others, sometimes without knowing all the contents—including the defamatory content—and sometimes without any opportunity to ascertain, in advance, that any defamatory matter was to be included in the matter published. The question is to what extent should one who is in the business of making available to the general public what another writes or says be subject to liability for the defamatory matter that was published.[11] P5. Courts have even held that a defendant can be deemed to have “adopted”—and thus to have published—statements posted by others on property within the defendant’s control if the defendant unreasonably fails to remove the statements despite having notice of their existence and a reasonable opportunity to take action.[12] P6. Participants in the communication of a statement other than the original author can be divided into three categories: (1) Those ordinarily responsible for the approval of content—such as book, newspaper, or magazine publishers—are primary publishers, sometimes referred to simply as publishers.[13] Primary publishers are deemed to have intended to publish the statements of the underlying author, and are treated like the underlying author because of their editorial control and opportunity to learn of the material to be published.[14] To be held liable, the primary publisher, like the original author, must also have the relevant degree of fault.[15] (2) Those who deliver content—such as libraries, bookstores, and news vendors—are secondary publishers, also called disseminators or distributors.[16] Although distributors, like primary publishers, are deemed to have intended to publish the statements of the underlying author, they can only be held liable for them if they knew or had reason to know of the defamatory content.[17] This knowledge requirement stems from First Amendment concerns that it would be unreasonable to require distributors to read all the materials they disseminate and might affect the amount and types of material they are willing to make available.[18] Like an original author and a primary publisher, it would appear that the distributor must also have the relevant degree of fault to be held liable.[19] (3) Those who merely provide the equipment and facilities the speaker used to communicate a statement—such as telephone companies—are not deemed to have published the statement for purposes of libel analysis because they cannot be said to have intended to publish the particular statement.[20] This is so even if they had notice of the purpose to which their equipment was being put.[21] P7. Thus, the term “publisher” can be used in both a broad and a narrow sense. In its broad sense, it refers to one who has sufficient responsibility in the communication of a statement to meet the publication element of libel, and includes both primary publishers and distributors. In its narrow sense, it refers to primary publishers as distinguished from distributors. This can lead to confusion, and was an issue in at least one on-line libel case.[22] For Lunney to demonstrate that Prodigy “published” the e-mail and bulletin board messages, then, he can show that Prodigy acted as a primary publisher of the messages by exercising sufficient editorial control over them, or acted as a distributor, perhaps by unreasonably failing to remove the statements despite having notice of their existence and a reasonable opportunity to take action, and thus adopting them.[23] 2. To a Third Party P8. Libel is about injury to reputation, and so relates to the opinion others have of the plaintiff.[24] Consequently, the plaintiff generally must show that the libelous statement was communicated to a person other than the plaintiff.[25] If a statement is posted in the woods and no one reads it, there is no libel. Thus, this element requires Lunney to demonstrate that someone else read the e-mail and bulletin board messages.[26] 3. A Statement of and Concerning the Plaintiff P9. Also stemming from the fact that libel is about injury to reputation, the statement must be about the plaintiff.[27] The statement may refer to the plaintiff indirectly, so long as at least one recipient reasonably understands it to pertain to the plaintiff.[28] The “of and concerning” element therefore requires Lunney to demonstrate that someone reasonably could have read into the e-mail and bulletin board messages a statement about him. 4. That is Defamatory P10. Again, because libel is about injury to reputation, the statement must be defamatory.[29] A statement usually is said to be defamatory if it reduces the esteem in which the plaintiff is held, or deters people from associating with the plaintiff.[30] Mere vulgarity or name-calling, however, usually cannot sustain a libel claim.[31] Form is not important: defamatory communication can occur directly or indirectly, by question or insinuation, on the face of the statement or by context, so long as the message conveys a defamatory meaning.[32] Nor is the intent of the publisher relevant: a statement is deemed defamatory so long as a third party reasonably construed it as so.[33] This element, then, requires Lunney to demonstrate that the e-mail and bulletin board messages reasonably can be construed as containing statements that harmed his reputation, and that they were not mere epithets. 5. False P11. The statement must not only be defamatory; it usually must be proved false by the plaintiff.[34] Out of a concern for reputation, falsity is presumed at common law and truth is an affirmative defense.[35] U.S. Supreme Court decisions have held, however, that the First Amendment requires a plaintiff to show that the statement is false in suits brought by public officials or public figures, and in suits brought by private figures against media defendants in cases involving statements on matters of public concern.[36] As Lunney’s case appears to be one by a private-figure plaintiff against a non-media defendant on a matter not of public concern, Lunney may not need to prove the statements false, although the statements’ falsity does not appear to have been questioned in the case in the sense that the parties seem to agree that Lunney did not send them. 6. And Unprivileged P12. The statement also must not be privileged. To ameliorate the chilling effect that the threat of defamation liability might have on free speech, certain privileges protect potentially defamatory but socially valuable statements.[37] Two categories of privileges exist at common law: (1) absolute privileges, and (2) conditional privileges, also called qualified privileges.[38] P13. Absolute privileges provide complete immunity from libel liability based on the status of the speaker and the context of the speech.[39] They have been applied, for example, to participants in judicial, legislative, and administrative proceedings; to executive and administrative officers acting within the scope of their duties; and to communications between husbands and wives.[40] P14. Qualified privileges provide limited immunity from libel liability based on the occasion for the speech.[41] Someone who reports to the authorities what appears to be a crime, for example, may be deemed immune from liability if the report turns out to have been in error.[42] To be protected by a qualified privilege, the individual must make the statement in good faith.[43] If the statement is made with what is called “common-law” malice, the immunity has been “‘abused’ and the privilege lost.”[44] The term “malice” is “a source of endless confusion.”[45] In the context of defeasance of a conditional privilege, it has been defined as “any wrongful motivation that is inappropriate given the reason for the existence of the privilege, including spite, ill will, hatred, or the intent to inflict harm.”[46] To complicate matters, courts sometimes refer to common-law malice as “malice,” “express malice,” “malice in fact,” and “actual malice.”[47] In New York, both actual malice—meaning knowledge or reckless disregard of the falsity of a statement—and common-law malice defeat a qualified privilege.[48] As we will see, issues concerning absolute and qualified privileges will arise in Lunney’s case. 7. With Some Degree of Fault by the Defendant P15. Because of concern for reputation, one who intentionally publishes defamatory material is held liable at common law regardless whether the defendant is negligent regarding the false and defamatory nature of the communication.[49] To ameliorate the severity of such a rule with interests in free speech, the common law allows a defendant to raise the truth of the statement, or certain privileges, as defenses.[50] In a series of cases starting in 1964, however, the U.S. Supreme Court held that the First Amendment often requires a plaintiff to prove fault regarding the falsity of the defamatory statement.[51] In cases brought by public-official or public-figure plaintiffs, for example, media defendants may not be found liable absent a finding of actual malice, which is knowledge or reckless disregard of the falsity of the statement.[52] In cases brought by private-figure plaintiffs involving statements on matters of public concern, media defendants may not be found liable absent a finding of some degree of fault, although it need not rise to the level of actual malice.[53] P16. The U.S. Supreme Court has not addressed whether the actual-malice standard applies to suits by public officials or public figures against non-media defendants.[54] Nor has the Court addressed whether private-figure plaintiffs must demonstrate fault in cases about matters not of public concern, or in cases against non-media defendants.[55] The First Amendment requirements act as a floor, however, and state legislatures and courts may impose higher standards of fault than the Federal Constitution mandates in a particular situation. New York courts require private- figure plaintiffs in cases involving media defendants and statements on matters “‘arguably within the sphere of legitimate public concern’” to demonstrate “‘by a preponderance of the evidence, that the publisher acted in a grossly irresponsible manner without due consideration of the standards of information gathering and dissemination ordinarily followed by responsible parties.’”[56] New York courts apply the same test to cases brought by private-figure plaintiffs against non- media defendants involving statements on matters of public concern “where the test’s terminology permits its application.”[57] New York courts have yet to definitively rule, however, on the amount of fault a private-figure plaintiff must demonstrate in cases involving statements on matters not of public concern, which may be due in part to the courts’ willingness to deem almost any statement as “arguably within the sphere of legitimate public concern.”[58] As Lunney’s case could conceivably be deemed a matter involving a private-figure plaintiff, a non-media defendant, and a statement on a matter not of public concern, the standard of fault he is required to show may be low. 8. That Caused the Plaintiff Injury P17. At common law, a libelous statement must harm the plaintiff to be actionable, but damage to reputation is often presumed.[59] Most jurisdictions now require the plaintiff to show actual financial loss, called special damages, except in cases of libel per se.[60] “No concept in the law of defamation has created more confusion,” largely because, in the context of slander, “per se” refers to four specific categories of defamatory statements that do not require proof of special damages: allegations of crime, allegations that would harm another’s livelihood, allegations of loathsome disease, and allegations of sexual (mis)conduct.[61] In the context of libel, on the other hand, a communication is libelous per se if its defamatory meaning is apparent on the face of the statement, although some jurisdictions also include the four slander per se categories.[62] If the defamatory character is apparent only with knowledge of other circumstances, the communication is considered libel per quod, and requires proof of special damages.[63] New York has not been exempt from the general confusion surrounding libel per se.[64] “[T]he best guess remains that New York still adheres to the per se/per quod approach .... The issue cannot be settled, however, until a case squarely raising it is decided by the New York Court of Appeals.”[65] P18. The U.S. Supreme Court has held that the First Amendment prohibits the presumption of damages against a media defendant in cases involving statements on matters of public concern, absent a finding of actual malice.[66] Damages may constitutionally be presumed even absent actual malice, however, in cases involving private plaintiffs on matters not of public concern, and possibly in cases involving non-media defendants.[67] In light of the uncertainty in New York regarding the application of the fault requirement in cases involving non-media defendants or matters not of public concern,[68] and the application of the libel per se rules, it is unclear whether Lunney must demonstrate special damages. Because his case may be deemed one that involves a private-figure plaintiff, a non-media defendant, and a matter not of public concern, he may not need to prove special damages if he can demonstrate that the e-mail and bulletin board messages are libelous per se. B. Libel Online 1. Cubby, Inc. v. CompuServe, Inc. P19. In one of the first significant on-line libel cases, 1991’s Cubby, Inc. v. CompuServe, Inc., the U.S. District Court for the Southern District of New York applied traditional libel analysis.[69] Cubby had sued CompuServe for libel in a diversity action based on statements in Rumorville USA, a daily newsletter about broadcast journalism that CompuServe carried in its Journalism Forum.[70] Cubby, which had developed a rival electronic publication called Skuttlebut, alleged that Rumorville contained defamatory statements regarding Skuttlebut’s reporting credentials and techniques.[71] CompuServe sought summary judgment on the grounds that it was not a primary publisher but a distributor of the statements, and neither knew, nor had reason to know, of the allegedly defamatory statements.[72] P20. Analogizing CompuServe to an electronic library, the court observed that CompuServe had no more editorial control over Rumorville than public libraries, bookstores, or newsstands have over the publications they carry.[73] CompuServe had contracted with an independent company, Cameron Communications Inc, to “manage, review, create, delete, edit and otherwise control the contents” of the Journalism Forum.[74] Cameron Communications had, in turn, contracted with the publisher of Rumorville, Don Fitzpatrick Associates, to include it in the forum.[75] CompuServe had “no employment, contractual, or other direct relationship with” Don Fitzpatrick Associates.[76] And although the contract between CompuServe and Cameron Communications stated that Cameron would operate the forum “in accordance with editorial and technical standards and conventions of style as established by CompuServe,” the contract between Cameron Communications and Don Fitzpatrick Associates stated that Don Fitzpatrick Associates “accepts total responsibility for the contents” of Rumorville.[77] The court also found no agency relationship between CompuServe and either Cameron Communications or Don Fitzpatrick Associates.[78] P21. Reasoning that “[a] computerized database is the functional equivalent of a more traditional news vendor,” the court concluded that the free flow of information required application of the same standard of liability as would apply to a public library, bookstore, or newsstand.[79] For CompuServe to examine all the publications it makes available, the court stated, would be just as infeasible as it would for any other distributor.[80] Consequently, the court applied the distributor liability standard and examined whether CompuServe knew or had reason to know of the allegedly defamatory statements.[81] CompuServe had claimed it had no notice of any complaints about the newsletter or Don Fitzpatrick Associates.[82] CompuServe also had contended that it had neither knowledge nor reason to know of the allegedly defamatory statements, particularly in light of the number of publications it carries and the speed with which Don Fitzpatrick Associates made its publication accessible to CompuServe customers.[83] Because Cubby did not set forth specific facts demonstrating a genuine issue regarding whether CompuServe knew or had reason to know of Rumorville’s contents, the court granted CompuServe summary judgment on the libel claim.[84] Thus, if one were to scrutinize Lunney’s libel claims under a Cubby- type analysis, it would appear that he must show either that Prodigy exercised sufficient editorial control over the e-mail and bulletin board messages to be a primary publisher, or that it acted as a distributor and knew or should have known that the messages contained defamatory content. 2. Stratton Oakmont, Inc. v. Prodigy Servs. Co. P22. Stratton Oakmont, Inc. v. Prodigy Servs. Co. presents an example of a case in which the existence of editorial control in the on-line context cast a service provider as primary publisher rather than a distributor. Picking up where Cubby, Inc. v. CompuServe, Inc. left off, the N.Y. Supreme Court stated in Stratton that it was in general agreement that “[c]omputer bulletin boards should generally be regarded in the same context as bookstores, libraries and network affiliates.”[85] The court deemed Prodigy to be the primary publisher of statements posted on one of its bulletin boards, nonetheless, on the grounds that Prodigy exposed itself to greater liability by exercising editorial control over the content.[86] P23. The case arose when an unidentified individual posted statements on Prodigy’s “Money Talk” bulletin board in October 1994 accusing the Stratton Oakmont investment-banking firm of criminal securities fraud.[87] Stratton sued Prodigy for libel per se,[88] and sought summary judgment on the issue of whether Prodigy could be considered a publisher of the statements.[89] The court granted Stratton’s summary judgment motion for two reasons.[90] P24. First, Prodigy had marketed itself as a family- oriented on-line service that controlled the content of its bulletin boards, differentiating itself from other on-line services and explicitly likening itself to a newspaper.[91] Second, pursuant to policies that the court said might chill on-line expression, Prodigy used software that automatically screened for offensive language before posting messages,[92] and contracted with “Board Leaders” to enforce certain content guidelines.[93] Using technology and personnel to remove certain material from bulletin boards, the court said, amounted to decision-making over content.[94] The court acknowledged that Prodigy’s control was not total, but concluded nonetheless that Prodigy had “uniquely arrogated to itself the role of determining what [was] proper for its members to post and read on its bulletin boards.”[95] Consequently, the court concluded that Prodigy was a primary publisher of the bulletin board messages, not a distributor.[96] Thus, under a Stratton analysis, Cubby does not control Lunney’s claims if Prodigy had editorial control over the e-mail and bulletin board messages. P25. Note that in Cubby, CompuServe required Cameron Communications to apply CompuServe’s standards,[97] much like Prodigy hired board leaders to apply Prodigy’s standards. Unlike the situation in Cubby, however, the Stratton court found an agency relationship between the Board Leaders and Prodigy.[98] Cubby also involved an article in an on-line forum, rather than a bulletin board posting, and did not involve automated screening software. These differences may be sufficient to reconcile the cases, and at least one author finds both cases to be consistent applications of traditional libel analysis.[99] P26. In addressing concerns that its ruling could deter service providers from monitoring the content on their services, the court argued that the market might compensate service providers for exercising editorial control and exposing themselves to liability.[100] The court also observed that passage of the Communications Decency Act, several versions of which were then pending in Congress, might make its decision moot.[101] 3. The Communications Decency Act P27. Congress did, indeed, pass the Communications Decency Act, which became effective February 8, 1996.[102] The CDA added section 230 to the Communications Act of 1934, and took a decidedly different tack toward libel on-line. Apparently unconvinced that market forces would offset any disincentive toward editorial control that increased exposure to liability might bring, Congress overruled Stratton with section 230(c), labeled “Protection for ‘good samaritan’ blocking and screening of offensive material.”[103] Under section 230(c)(1), “[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.”[104] P28. Congress also specifically sought to immunize interactive computer services from liability for attempting to restrict access to objectionable material. Thus, section 230(c)(2) states that: [n]o provider or user of an interactive computer service shall be held liable on account of— (A) any action voluntarily taken in good faith to restrict access to or availability of material that the provider or user considers to be obscene, lewd, lascivious, filthy, excessively violent, harassing, or otherwise objectionable, whether or not such material is constitutionally protected; or (B) any action taken to enable or make available to information content providers or others the technical means to restrict access to [information provided by another information content provider].[105] P29. The language of the statute indicates that Congress had at least two motives behind its actions. First, Congress sought to help the embryonic on-line industry flourish.[106] Second, Congress sought to encourage interactive computer services to screen the content they provide and to make available to their users means of limiting their exposure to certain types of material.[107] In an attempt to ensure that Congress’ policy judgements become the law of the land, section 230(e)(3) states that “[n]o cause of action may be brought and no liability may be imposed under any State or local law that is inconsistent with this section.”[108] P30. Because the immunity conferred by section 230(c)(1) is based on the defendant’s status as an interactive computer service, and appears to apply even to publications made in bad faith,[109] it can be seen as an absolute privilege against liability for third- party speech.[110] Section 230(c)(2)(B) could also be viewed as conferring an absolute privilege. Because section 230(c)(2)(A) applies only to actions “voluntarily taken in good faith,” however, it appears to create a qualified privilege. P31. Section 230 potentially presents a significant stumbling block to Lunney’s claims. Sub-subparagraph (c)(2)(A) implies that Lunney cannot rely on the use by Prodigy of screening software or Board Leaders to impose primary publisher liability on Prodigy, so long as Prodigy made such use, if at all, in good faith. Of greater significance, however, is subparagraph (c)(1), which suggests that Prodigy may be absolutely immune from liability for the e-mail and bulletin board messages on the grounds that Prodigy cannot be deemed to have published third-party speech. Depending upon whether one construes the subparagraph as invoking the broad or narrow meaning of “publisher,” however, the subparagraph could be viewed as protecting interactive computer services only from primary publisher liability, and not from distributor liability.[111] As we will see, Zeran v. America Online Inc. and Blumenthal v. Drudge answer that question. They raise other questions, however, regarding retroactive application of the CDA, and the difference between holding an on-line service liable as an interactive computer service and as an information content provider. 4. Zeran v. America Online Inc. P32. By the time Congress passed the CDA, the circumstances that would lead to the first significant interpretation of section 230 were already in motion. In Zeran v. America Online Inc., the U.S. Court of Appeals for the Fourth Circuit concluded that section 230 immunizes interactive computer services from the primary publisher liability applicable to print and broadcast media, as well as the distributor liability applicable to entities such as bookstores and libraries.[112] P33. Six days after the April 19, 1995, Oklahoma City bombing of the Alfred P. Murrah Federal Building, someone posted a message on an AOL electronic bulletin board.[113] The message advertised t-shirts with what many people found to be offensive slogans relating to the bombing.[114] The bulletin board message indicated that people could purchase the shirts from “Ken,” and listed the home telephone number of Ken Zeran.[115] Inundated with hateful telephone calls, some of which included death threats, Zeran called AOL to report that he had been the victim of a hoax.[116] A company representative told him that AOL would remove the message, but would not post a retraction as a matter of policy.[117] The first advertisement was apparently removed by the next day, but similar postings persisted for another five days, as did the threatening calls.[118] Zeran continued to contact AOL, and company representatives told him that AOL would soon close the account responsible for the postings.[119] P34. Zeran filed suit April 23, 1996, seeking to hold AOL liable for the defamatory postings of the unidentified third party.[120] Zeran argued that once he notified AOL of the hoax, the company had an obligation to promptly remove the defamatory messages, to post public retractions on its boards, and to screen for similar postings.[121] The U.S. District Court for the Eastern District of Virginia granted AOL’s motion for judgment on the pleadings, holding that the CDA provided AOL an affirmative defense barring Zeran’s claims.[122] Zeran appealed.[123] i. Publisher and Distributor Liability P35. The U.S. Court of Appeals for the Fourth Circuit affirmed the lower court decision, holding that section 230(c)(1) immunized AOL from liability stemming from the fake advertisements.[124] The court stated that section 230(c)(1) bars claims that seek to hold service providers liable for exercising the traditional editorial functions of publisher, “such as deciding whether to publish, withdraw, postpone or alter content.”[125] Zeran argued, however, that Stratton and Cubby recognize a distinction between “publishers” and “distributors.”[126] He claimed that section 230 immunizes service providers only from publisher liability, and that by informing AOL of the hoax, he had given AOL sufficient notice of the defamatory statements for purposes of assessing distributor liability.[127] The court disagreed.[128] P36. “Publication,” the court said, covers not only an author’s intentional inclusion of information, but also the negligent communication of a defamatory statement and the failure to omit such a statement when relaying information to another.[129] “‘[E]very one who takes part in the publication ... is charged with publication,’” the court explained, even distributors.[130] Distributor liability, the court said, is merely a subset of publisher liability.[131] Although different standards of liability may apply to different types of publishers, according to the court all fall within the scope of section 230.[132] AOL, the court found, was a publisher within the meaning of section 230, and thus within the immunity it conferred.[133] P37. Furthermore, the court reasoned, holding service providers culpable based on notice would defeat section 230’s goal of encouraging them to regulate their content.[134] Distributor liability would subject service providers to potential litigation whenever they receive notice of a possibly defamatory statement, requiring them to quickly assess whether the content is indeed libelous and to decide whether to permit publication.[135] The court observed that “[a]lthough this might be feasible for the traditional print publisher, the sheer number of postings on interactive computer services would create an impossible burden in the Internet context.”[136] Thus, the court concluded, distributor liability—like the primary publisher liability imposed in Stratton—would discourage service providers from monitoring content for fear of learning of potentially defamatory material and subjecting themselves to liability.[137] Consequently, if section 230 applies to Lunney’s claims, Zeran suggests that Lunney cannot hold Prodigy liable as either a primary publisher or a distributor of the e-mail and bulletin board messages. Other aspects of Zeran, however, leave open the possibility that section 230 will not apply to Lunney’s case. ii. Retroactive Application of the CDA P38. Zeran had also argued that the February 8, 1996, effective date of the CDA and the presumption against retroactive application of statutes made section 230 irrelevant to the April 1995 circumstances underlying his case,[138] but to no avail. The district court held that the CDA applies to suits filed after its enactment, regardless when the underlying circumstances arose,[139] and the appeals court agreed. Section 230(e)(3) states that “[n]o cause of action may be brought and no liability may be imposed under any State or local law that is inconsistent with this section.”[140] According to the court, “retroactivity concerns arise when a statute applies to conduct predating its enactment.”[141] The court noted that section 230 prohibits bringing certain causes of action, and Zeran did not bring his cause of action until April 23, 1996, two months after section 230 became effective.[142] The court concluded, therefore, that applying section 230 to the Zeran case was prospective.[143] P39. Even if analyzing the Zeran case under section 230 could be considered retroactive application of a federal statute, the court concluded that the U.S. Supreme Court’s decision in Landgraf v. USI Film Products required such retroactive application.[144] To determine whether a statute applies retroactively under Landgraf, a court must “first ‘determine whether Congress has expressly prescribed the statute’s proper reach.’”[145] According to the Zeran court, section 230’s “no cause of action may be brought” language clearly expresses a congressional intent that the statute apply to any complaint filed after the CDA’s enactment, regardless when the underlying circumstances arose.[146] The Zeran court saw in section 230 a congressional determination “that free speech on the Internet and self-regulation of offensive speech were so important that §230 should be given immediate, comprehensive effect.”[147] P40. AOL had made an even broader argument before the district court below, claiming that the CDA applies to all causes of action pending after its enactment, even if filed before its effective date.[148] AOL based its claim on the fact that section 230(e)(3) states that “[n]o cause of action may be brought and no liability may be imposed” that is inconsistent with section 230.[149] AOL argued that no statutory provision should be read to render other provisions of the statute superfluous, and concluded that section 230(e)(3)’s liability clause expresses a congressional intent to apply the CDA to pending actions.[150] The district court chose not to address the issue, stating that although “AOL’s argument is not without considerable force, … this close question is immaterial to determining the retroactive application of the CDA to Zeran’s cause of action, which was brought two months after the enactment of the CDA.”[151] Thus, in light of Zeran, the fact that the circumstances underlying Lunney’s claims occurred before enactment of the CDA does not prevent Prodigy from raising section 230 as a defense. The fact that Lunney filed his initial claims in December 1994, however, remains a potential obstacle that Prodigy must overcome. 5. Blumenthal v. Drudge P41. AOL was a party in another on-line libel case decided just five months following the Fourth Circuit’s opinion in Zeran v. America Online Inc. In Blumenthal v. Drudge, the U.S. District Court for the District of Columbia relied on Zeran to grant AOL summary judgment in a libel case over statements by rumor reporter Matt Drudge, but not before criticizing section 230 as inequitable.[152] The suit involved the Drudge Report, a political rumor column that Matt Drudge made available on his web site and by e-mail, and that AOL posted on its service under a licensing agreement with Drudge.[153] In the August 10, 1997, edition, Drudge had written about allegations of past spousal abuse by then recently hired White House aide Sidney Blumenthal.[154] Drudge issued a retraction on August 12, one day after receiving a letter from Blumenthal’s lawyer.[155] AOL also posted the retraction and later removed the August 10 Drudge Report from its archive of previous editions.[156] The Blumenthals sued Drudge and AOL, nonetheless, and AOL moved for summary judgment.[157] P42. The court observed that section 230 effectively immunizes interactive computer services from tort liability for material they transmit but do not create.[158] The court construed section 230(c)(1) as insulating interactive computer services from liability for any “failure to edit, withhold or restrict access to offensive material” they carry.[159] Thus it was irrelevant, the court said, that the Washington Post might be held liable if, like AOL, it had “‘publish[ed] Drudge’s story without doing anything whatsoever to edit, verify, or even read it (despite knowing what Drudge did for a living and how he did it).’”[160] P43. The Blumenthals conceded that AOL was a “provider … of an interactive computer service,” and that if it acted exclusively as such, it could not be held liable for making the Drudge Report available.[161] The Blumenthals sought, however, to hold AOL liable on the grounds that it was a joint creator of the Drudge Report, and thus an “information content provider.”[162] They argued that Drudge was not just an anonymous individual that used AOL to send a message over the Internet.[163] Rather, AOL had contracted with Drudge; provided his sole, consistent source of income; promoted the scandalous nature of the Drudge Report to existing and potential subscribers; and reserved the right to remove, or instruct Drudge to remove, content from the Drudge Report.[164] P44. This raises an interesting question: at what point does a “provider or user of an interactive computer service” exercise sufficient editorial control over the content on a service to become the underlying “information content provider” of that content?[165] If the service provider was so intertwined with the creation of the content that it could be deemed the underlying information content provider, section 230 would not shield it from liability because the content would be its own, not that of a third party.[166] Thus, analysis of a service provider’s editorial control over content may be relevant even under a section 230 framework because that control may convey upon the service provider the status of information content provider. P45. In the Drudge case, however, the court noted that the Blumenthals themselves conceded that AOL played neither an editorial nor a substantive role in creating the Drudge Report.[167] The court also found that Drudge was neither an employee nor agent of AOL.[168] The court concluded, therefore, that AOL was merely an interactive computer service that carried the Drudge Report, and could not be treated as a publisher or speaker of the statements at issue in light of section 230(c)(1).[169] “If it were writing on a clean slate,” the court said, it would agree with the Blumenthals that AOL should not be allowed “to tout someone as a gossip columnist or rumor monger” and then claim immunity when that person defames someone:[170] AOL has certain editorial rights with respect to the content provided by Drudge and disseminated by AOL, including the right to require changes in content and to remove it; and it has affirmatively promoted Drudge as a new source of unverified instant gossip on AOL. Yet it takes no responsibility for any damage he may cause. AOL is not a passive conduit like the telephone company, a common carrier with no control and therefore no responsibility for what is said over the telephone wires. Because it has the right to exercise editorial control over those with whom it contracts and whose words it disseminates, it would seem only fair to hold AOL to the liability standards applied to a publisher or, at least, like a book store owner or library, to the liability standards applied to a distributor.[171] P46. The court pointed out, however, that the slate was not clean: Congress has made a different policy choice by providing immunity even where the interactive service provider has an active, even aggressive role in making available content prepared by others. In some sort of tacit quid pro quo arrangement with the service provider community, Congress has conferred immunity from tort liability as an incentive to Internet service providers to self-police the Internet for obscenity and other offensive material, even where the self-policing is unsuccessful or not even attempted.[172] P47. The court also noted that, in light of Zeran, categorizing AOL as a distributor of the Drudge Report rather than a publisher would make no difference.[173] Thus, although it appeared to the court that AOL had taken “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,”[174] the court held that section 230 immunized AOL from suit.[175] Consequently, the court granted AOL’s motion for summary judgment.[176] P48. In the wake of Drudge, then, Lunney would seem to have quite an uphill battle if section 230 applies to his claims. Not only does Drudge reaffirm that section 230 immunizes interactive computer services from distributor as well as primary publisher liability, but it also suggests that such immunity is absolute, and is not subject to defeasance even if Prodigy acted in bad faith. The case raises an additional argument, however, for plaintiffs to use in warding off application of section 230: that the service provider is liable not as an interactive computer service, but as an information content provider. III. The Lunney Case P49. Despite the Zeran and Drudge decisions, the Appellate Division of the N.Y. Supreme Court[177] chose not to apply section 230 of the Communications Act in its December 1998 Lunney v. Prodigy Servs. Co. libel decision. Instead, the court examined Lunney’s complaint in light of the elements of libel, analogized Prodigy to telephone and telegraph companies, and extended to service providers a qualified, state common-law privilege traditionally applied to common carriers.[178] A. The Suit P50. On September 9, 1994, someone sent a Boy Scout troop leader a sexually explicit and menacing e-mail message.[179] Based on the text of the message and the Prodigy account from which it was sent, the message appeared to come from prospective Eagle Scout Alex Lunney, who was then 15 years old.[180] The Boy Scout leader reported the matter to the Bronxville police and to Lunney’s Scout Master.[181] The Scout Master confronted Lunney at home in front of his mother.[182] Lunney denied sending the message.[183] Apparently neither the police nor the Boy Scouts ever took action against Lunney, suggesting that they believed him.[184] P51. The matter did not end there, however. Lunney received a letter from Prodigy dated September 14, 1994, informing him that Prodigy had suspended his account for transmission of “abusive, obscene, and sexually explicit material.”[185] Lunney responded that he had never opened a Prodigy account.[186] Prodigy apologized the following month, and told Lunney that it had closed a number of fraudulent accounts that had been opened in his name.[187] No one has been identified as the author of the e-mail or bulletin board messages.[188] P52. In December 1994, Lunney’s father filed suit in his son’s name in the N.Y. Supreme Court, accusing Prodigy of libel per se based on the e-mail message.[189] Prodigy moved for summary judgment, which the court denied in July 1996 “without prejudice to renewal on completion of discovery.”[190] The court denied the motion on the grounds that it could not rule as a matter of law that Prodigy was not a publisher in light of provisions in Prodigy’s service agreement that appeared to reserve to it a certain amount of control over content.[191] P53. Prodigy renewed its summary judgment motion after Lunney completed discovery.[192] Having uncovered in discovery that two electronic bulletin board messages harassing other individuals were posted in his name just days before the e-mail message,[193] Lunney opposed Prodigy’s summary judgment motion and moved in April 1997 to amend his complaint to include the bulletin board messages.[194] The court granted Lunney’s request in a July 1997 order over Prodigy’s objection that the new claims were defective and barred by section 230.[195] That order also denied Prodigy’s renewed summary judgment motion, but said that Prodigy could file its motion again after conclusion of discovery on the new allegations.[196] P54. After Lunney filed his amended complaint and indicated he did not need additional discovery, Prodigy renewed its summary judgment motion yet again, incorporating by reference its submission in support of the earlier summary judgment motion.[197] The court denied the motion in January 1998 on the grounds that it was not the court’s responsibility to requisition the file to review the prior submission, and that the prior submission was outdated and related to the original complaint.[198] Prodigy appealed both the July 1997 and January 1998 orders.[199] The appeals, which were consolidated, argued that the lower court erred by allowing Lunney to amend his complaint, and by denying Prodigy’s motions for summary judgment.[200] B. The Appellate Division Appeal P55. In its brief before the Appellate Division, Prodigy cited the distributor liability analysis of Cubby, Inc. v. CompuServe, Inc., and argued that Lunney had not shown that Prodigy either knew, or had reason to know, of the allegedly defamatory statements.[201] Prodigy also argued that even if it did know of the statements, it could not be deemed to have published the messages transmitted over its lines under Anderson v. New York Tel. Co., a case involving New York common- law principles traditionally applied to telephone and telegraph companies.[202] In any event, Prodigy claimed, the statements were not defamatory and, although they purported to be from Lunney, were not “of and concerning” him.[203] At worst, Prodigy argued, they could lead someone to infer that Lunney “is a foul-mouthed teenager.”[204] Even if such an accusation were made directly concerning Lunney, Prodigy argued, it would amount to nothing more than a non-actionable epithet.[205] Prodigy argued, moreover, that the statements did not constitute libel per se, and Lunney failed to show special damages.[206] P56. Prodigy also argued that the Communications Decency Act barred Lunney’s libel claims.[207] Prodigy made that argument even though, unlike in Zeran v. America Online Inc., the circumstances underlying Lunney’s claims and the date Lunney filed his initial complaint both occurred prior to the effectiveness of the CDA.[208] Prodigy contended that the allegedly libelous electronic bulletin board messages that Lunney raised in his April 1997 amended complaint did not relate back to the original claim of libel because the messages were separate, distinct statements posted on separate days, possibly even by different people.[209] Consequently, Prodigy cited the Zeran court’s retroactivity analysis to argue that section 230 applied to the electronic bulletin board messages, which Lunney raised after the effectiveness of the CDA.[210] Resurrecting the argument that AOL made in Zeran—and that the district court in Zeran chose not to reach—Prodigy also argued that section 230 prohibits courts from imposing libel liability for claims brought but not decided before effectiveness of the CDA.[211] Thus, Prodigy argued that the CDA applied to the e-mail message in Lunney’s original complaint.[212] Such a conclusion, Prodigy said, is consistent with the Zeran court’s statement that “Congress decided that free speech on the Internet and self-regulation of offensive speech were so important that § 230 should be given immediate, comprehensive effect.”[213] P57. Lunney, on the other hand, noted that Prodigy had already been deemed a publisher in Stratton Oakmont Inc. v. Prodigy Servs. Co.[214] Lunney argued that Cubby was not controlling because this case involved e- mail and bulletin board messages, rather than the distribution of “news.”[215] Lunney distinguished the Anderson case on the grounds that a public utility such as a telephone company does not possess the amount of editorial control that Prodigy reserves for itself under its member service agreement.[216] With regard to proof of damages, Lunney argued that none was necessary because the messages suggesting that Lunney was threatening to commit rape and murder constituted libel per se.[217] P58. Turning to the retroactivity issue, Lunney argued that the Zeran decision, and thus the CDA, did not apply because he filed his claim before the statute became effective.[218] Lunney argued, moreover, that statutes are presumed to apply only prospectively absent a clearly expressed congressional intent to the contrary.[219] The “no liability shall be imposed” language of section 230, Lunney contended, did not constitute a clear expression of such congressional intent.[220] C. The Appellate Division's Analysis P59. The Appellate Division reversed the lower court and granted Prodigy summary judgment, concluding that e-mail and bulletin board messages did not appear defamatory.[221] The court reasoned that although the messages could lead a reader to infer that Lunney, as the supposed author, was a “foul-mouthed teenager,” the messages were not “of or concerning” Lunney.[222] “At most,” the court found, “one could read into the e-mail message in question a statement of fact that [Lunney] is a bully who has threatened to sodomize a scout leader’s sons.”[223] Moreover, even assuming such a statement was defamatory, the court concluded that Prodigy could not be held legally responsible for it, or for the contents of the bulletin board postings, for two reasons.[224] 1. Prodigy Not a “Publisher” P60. First, the court reasoned, “Prodigy did not publish the statement.”[225] Under New York common law, “[h]e who furnishes the means of convenient circulation, knowing, or having reasonable cause to believe, that it is to be used for that purpose, if it is in fact so used, is guilty of aiding in the publication and becomes the instrument of the libeler.”[226] Relying on the N.Y. Court of Appeals’ decision in Anderson v. New York Telephone Co., however, the Lunney court said “that no potential for such liability exists, unless the defendant in question has some ‘editorial or at least participatory function’ in connection with the dissemination of the defamatory material.”[227] P61. In Anderson, a church bishop sued the telephone company because it leased equipment to an individual who used it to play allegedly defamatory recorded messages to anyone that called either of two telephone numbers.[228] The bishop brought the messages to the attention of the telephone company, but the company apparently took no action.[229] The trial court directed a verdict for the telephone company, concluding that it did not publish the recordings.[230] The Appellate Division reversed, holding that a telephone company should be liable when it transmits a recorded defamatory message with knowledge or reckless disregard of the message’s falsity.[231] The Court of Appeals, adopting the Appellate Division dissent of Justice Witmer, reinstated the trial court’s judgment.[232] The Court of Appeals reasoned that telephone companies—unlike newspaper and magazine publishers, or radio, television, and telegraph companies—have no control over the content they transmit and thus cannot be deemed to have published the messages.[233] The court noted, for example, that employees of a telegraph company prepare the messages the company sends, whereas the employees of a telephone company play no such role.[234] The court also noted that the telephone company is a public utility that may refuse service only for certain reasons, none of which include the defamatory nature of the transmission.[235] Thus, the Court of Appeals concluded that use of the telephone company’s equipment did not render it a publisher, even if it had notice of the purpose for that use.[236] P62. The Lunney court found the legal analysis in Anderson “as persuasive in the case of written text sent over telephone lines, which is the essential nature of ‘e-mail’ and bulletin board postings, as it was in the case of spoken or recorded messages sent over telephone lines ....”[237] Applying that analysis, and finding Prodigy’s role more akin to that of a telephone company than that of a telegraph company, the court concluded that Prodigy had no participatory function in the communication of the e-mail and bulletin board messages.[238] It made no difference that Prodigy had devised a way to automatically exclude certain vulgarity from messages sent through its network, the court said, because “application of any unintelligent automated word-exclusion program cannot be equated with editorial control.”[239] Absent in the case, the court observed, was evidence that any Prodigy employee approved the specific e-mail[240] or bulletin board messages at issue.[241] Thus, the court ruled, Prodigy could not be deemed to have published the e- mail or bulletin board messages, and so could not be held culpable for libel.[242] P63. Nor did the court believe that Stratton required otherwise. According to the court, Stratton would apply, if at all, only to the bulletin board messages involved in Lunney.[243] The court reasoned that services such as Prodigy cannot screen all the e-mail their subscribers send, no matter how much editorial control the service providers exercise over their bulletin boards.[244] The court also observed that Prodigy had abandoned the editorial control procedures underlying Stratton in January 1994, making the factual context of Lunney different.[245] The court, relying again on Anderson, said that a telecommunications company can be considered to have published a message only when it has actually participated in its transmission.[246] Extending that logic, the court reasoned that a telecommunications company’s participation in the transmission of some messages cannot expose it to liability for the messages it did not exert control over.[247] Consequently, the court found irrelevant Prodigy’s ability to exclude vulgarity from some messages, as no evidence suggested that Prodigy had exercised such editorial control over the messages involved in Lunney.[248] 2. The Qualified, State Common-Law Privilege P64. Second, the court concluded that Prodigy could not be held culpable even if it did “publish” the statement.[249] According to the court, Anderson concluded not only that a telephone company is not a publisher for purposes of libel analysis, but also that even if it were, it would be shielded by a qualified, common-law privilege traditionally afforded telegraph companies.[250] This qualified privilege developed to ameliorate the impact of treating telegraph companies as publishers of their customers’ messages.[251] The Lunney court said that this privilege shields a telegraph company from liability for its customers’ defamatory statements absent a showing of actual malice, meaning knowledge of the message’s falsity.[252] As Justice Witmer’s Appellate Division dissent explained, the privilege recognizes that telecommunications subscribers have due process and First Amendment rights to service that the public utility may not “lightly abridge” by refusing transmission.[253] Thus, the privilege affords the utility some protection from libel suits, as well as invasion of privacy claims and the administrative burdens it would otherwise face if it needed to monitor all its transmissions.[254] P65. Concluding that “[e]-mail is, in substance, nothing but an updated version of the telegraph,” the Lunney court extended the qualified privilege to service providers.[255] Apparently finding no evidence of actual malice, the court concluded that the privilege protected Prodigy from liability even if it did “publish” the statement, providing yet another reason why Stratton was not dispositive.[256] 3. The Court's Assessment of On-line Libel Law P66. The court also took the Lunney case as an opportunity to express its disapproval of the Stratton Oakmont Inc. v. Prodigy Servs. Co. rationale for what it called a lack of fairness.[257] The Stratton decision “punished” Prodigy for inadequately exercising editorial control that it had no legal obligation to perform, the Lunney court stated.[258] Such a holding, the Lunney court criticized, encourages service providers to maintain mere “distributor” status by ignoring the substance of messages posted to their bulletin boards.[259] Thus, the Lunney court argued, Stratton discourages the conduct that the plaintiff in that case sought to promote.[260] P67. The court also maintained that applying Anderson as it did placed it in “complete harmony” with case law — such as Cubby, Zeran, and Drudge — that holds that service providers are not liable for damages when third parties use their services to transmit offensive or libelous messages.[261] Again placing its decision in a broader context, the Lunney court concluded that its application of the qualified privilege in Anderson put its decision “in complete harmony with ... the Communications Decency Act.”[262] Consequently, the court said that it need “not decide what is thus the essentially academic question of whether this Federal statute would apply to all or part of the allegations of the plaintiff’s amended complaint.”[263] IV. Discussion P68. The most striking feature of the Lunney opinion is the Appellate Division’s decision not to apply section 230. The Lunney court maintains that the way it disposes of the case places it “in complete harmony” with decisions such as Cubby, Zeran, and Drudge, and eliminates any need to address section 230.[264] But if the cases are consistent, why not apply section 230 as construed by Zeran and Drudge, or traditional libel law as construed by Cubby and Stratton? The court seems more than willing to provide multiple alternative rationales[265]— why not include these? P69. The Appellate Division most likely avoided section 230 so that it would not need to address the retroactivity question.[266] The Lunney court also may have been reluctant to apply the CDA because, as the Drudge court pointed out, section 230 appears to immunize interactive computer services from liability for third-party statements regardless whether the service providers act responsibly.[267] Section 230 is, nonetheless, the law of the land, and courts are not free to substitute their own policy judgments absent a finding that section 230 is either unlawful or inapplicable, a finding that the court chose to sidestep. P70. As for Stratton, and to a lesser degree Cubby, the court may not have wanted to apply traditional libel analysis in light of Congress’ decision with section 230 to abrogate application of the distributor/publisher framework to interactive computer services.[268] The basis for Congress’ decision was its criticism, echoed by the Lunney court, that traditional libel analysis discourages service providers from monitoring the content they provide.[269] The Lunney court’s own rationale, however, is susceptible to the same criticism because it leaves open the possibility that some degree of editorial control will confer publisher status upon service providers.[270] The court also may not have wanted to apply Cubby because that case focused on an on-line database, rather than e-mail or bulletin board messages, but this distinction hardly seems dispositive. Cubby simply applied traditional libel analysis, which examines the level of editorial control to determine the appropriate standard of liability. E-mail and electronic bulletin boards may raise different issues of control, but they are no less susceptible to traditional libel analysis.[271] P71. The Lunney court’s assertion that its decision is “in complete harmony” with cases such as Zeran, Drudge, and Cubby is also an overstatement.[272] Although the Lunney court claims that the outcome is consistent in the sense that it finds that Prodigy is not culpable for the e-mail and bulletin board messages, the rationales the court uses to reach that outcome are incompatible with analysis under section 230 and traditional libel law. P72. Under section 230, “[n]o cause of action may be brought and no liability may be imposed” that is inconsistent with its prohibition against holding an interactive computer service liable as the publisher of third-party content.[273] By engaging in analyses concerning the elements of libel and qualified privileges, the Lunney court implicitly considers such a claim and entertains the possibility that liability may be imposed.[274] That the court reached a similar result as it would have under section 230 by finding no liability does not change the fact that, in different circumstances, its rationale could produce a finding that a service provider is the publisher of, and liable for, the defamatory statements of a third party. If section 230 applies to Lunney’s case, any attempt to hold a service provider culpable as a publisher of third-party content should be dismissed for failure to state a cognizable or compensable claim.[275] Thus, “the outcome in this case” is not “in complete harmony with ... the Communications Decency Act.”[276] P73. Nor does the court’s “application of the holding of the Court of Appeals in Anderson” result in it “being in complete harmony with” traditional libel law as represented by cases such as Cubby.[277] Under traditional libel analysis, the court would examine Prodigy’s editorial control to determine whether it was acting as a distributor or a primary publisher, and apply the applicable standard of fault.[278] The Lunney court does consider whether Prodigy had an “editorial or at least participatory function” concerning the e- mail and bulletin board messages. The court does so, however, not to determine whether Prodigy should be subject to distributor or publisher liability, but to raise a state common-law theory applicable to telephone companies.[279] Invoking Anderson v. New York Telephone Company, the Lunney court analogizes service providers to telephone companies, distinguishes them from telegraph companies, and concludes that Prodigy cannot be culpable for libel because it did not publish the messages.[280] Still not content to conclude its analysis, the Lunney court analogizes service providers to telegraph companies, despite its earlier argument to the contrary.[281] It does so to argue that even if Prodigy is to be deemed to have published the messages, it should benefit from the qualified privilege that traditionally applied to telegraph companies.[282] P74. Compounding matters, the Lunney court misstates the common-carrier privilege.[283] The privilege, as printed in the Restatement of Torts and quoted by Justice Witmer in his Appellate Division dissent, states that: A public utility under a duty to transmit a message is conditionally privileged to do so, even though it knows the message to be defamatory, unless (a) The sender of the message is not privileged to send it, and (b) The agent who transmits the message knows or has reason to know that the sender is not so privileged.[284] P75. Thus, the question is not whether the defendant knows or has reason to know that the statement is false, but whether the defendant knows or has reason to know that the sender is not privileged to send it. P76. Even more problematic is the court’s reliance on flawed analogies between service providers and telephone and telegraph companies. It draws these analogies to invoke the publication and privilege analyses of Anderson. Both aspects of the Anderson decision, however, are predicated on the status of telephone and telegraph companies as common carriers,[285] which cannot refuse reasonable requests for service.[286] A service provider, on the other hand, “is not a passive conduit like the telephone company, a common carrier with no control and therefore no responsibility for what is said over the telephone wires.”[287] As the U.S. District Court for the District of Columbia pointed out in Blumenthal v. Drudge, a service provider “has the right to exercise editorial control over those with whom it contracts and whose words it disseminates.”[288] Service providers often maintain a right to refuse or restrict the use of their services, including e-mail, as evidenced by recent attempts to constrain mass electronic mailings, often referred to as “spam.” Prodigy has itself asserted in the past that it is not a common carrier, and has no duty to accept all transmissions.[289] More importantly, the Federal Communications Commission, which is charged with regulating the provision of interstate communications,[290] has repeatedly held that Internet-related service providers are not common carriers under the Communications Act of 1934, as amended.[291] P77. Among other things, the Communications Act authorizes the Commission to entertain complaints regarding common carriers,[292] as well as to regulate the way common carriers charge for their services,[293] the way they connect with each other’s networks,[294] the way they expand their facilities,[295] and the way they use data about their customers.[296] Congress and the Federal Communications Commission have been reluctant to treat Internet-related service providers as common carriers for fear that their regulation as such might stifle the development of the Internet. Congress, for example, has observed that “[t]he Internet and other interactive computer services have flourished, to the benefit of all Americans, with a minimum of government regulation,” and stated that “[i]t is the policy of the United States … to preserve the vibrant and competitive free market that presently exists for the Internet and other interactive computer services, unfettered by Federal or State regulation.”[297] The Commission expressed similar sentiments in its February 1999 declaratory ruling regarding carrier compensation for Internet-bound traffic. In that ruling, the Commission recognized an FCC “obligation and commitment … to foster and preserve the dynamic market for Internet-related services,” and emphasized “the strong federal interest in ensuring that regulation does nothing to impede the growth of the Internet—which has flourished to date under [the FCC’s] ‘hands off’ regulatory approach.”[298] Thus, the Lunney court’s analogy of service providers to common carriers may be as ill-advised as was Prodigy’s attempt to analogize itself to a newspaper, which led to unintended consequences in Stratton.[299] P78. Ironically, the Lunney court’s decision, including its analogy of service providers to common carries, is almost entirely dicta. Once the court holds that the e- mail and bulletin board messages were not “of or concerning” Lunney, his case fails to meet one of the essential elements of libel, entitling Prodigy to summary judgment.[300] If the court was not going to apply section 230 or traditional libel analysis under Cubby and Stratton, it could—and probably should—have stopped there. Doing so would have avoided the court’s publication and qualified privilege analyses, as well as its analogies to common carriers. P79. So what is Lunney’s legacy? The decision does diminish the significance that Stratton placed on automated editing systems, and suggests that a service provider cannot be held liable for an e-mail message it has not exercised editorial control over.[301] Lunney does not appear, however, to completely abandon Stratton. Despite the Lunney court’s criticism of the case and the case’s apparent abrogation by Congress, the court seems to say merely that Stratton is inapplicable to e-mail not edited by the service provider. In doing so, the court implies that Stratton could still be good law as applied to bulletin board and e-mail messages over which service providers exercise control. Perhaps more significantly, the court also extends to on-line services the qualified privilege traditionally applied to telegraph companies.[302] Much of this is academic, however, as there are likely few, if any, cases pending that are based on complaints filed before the CDA. V. Conclusion P80. Lunney has appealed, and both he and Prodigy have reiterated the libel-related arguments they made before the Appellate Division.[303] When the Court of Appeals[304] decides the case, which it heard October 13, 1999, it should determine whether application of section 230 to all or part of Lunney’s e-mail and bulletin board claims would be impermissibly retroactive.[305] Even if such application would not be impermissibly retroactive, section 230 still might not apply to one or both of those claims if Prodigy exercised sufficient editorial control over the e-mail or the bulletin board messages to be deemed an information content provider.[306] The Appellate Division appears to have found, however, that Prodigy did not examine the e-mail message, and abandoned its screening policies regarding bulletin board messages, making a finding of sufficient editorial control unlikely.[307] To the extent the Court of Appeals determines that section 230 does apply to the e-mail or bulletin board messages, it should affirm the summary judgment on the grounds that Prodigy is absolutely immune from libel liability for the publication of third-party speech, regardless of any bad faith on its part.[308] P81. If, on the other hand, the Court of Appeals concludes that section 230 does not apply, it should determine whether the Appellate Division is correct that the e-mail and bulletin board messages are not, as a matter of law, “of or concerning” Lunney. If the Court of Appeals concludes that the Appellate Division is correct, it should affirm the decision on the e-mail and bulletin board claims. At first blush, it would seem that the Appellate Division is correct, as the statements are not about Lunney, but purport to be by Lunney. This is not a clear cut question, however. As discussed above, the “of and concerning” element only requires Lunney to demonstrate that someone reasonably could have read into the e-mail and bulletin board messages a statement about him, either on the face of the messages or by inference.[309] Concluding that someone could not reasonably have read into the e-mail and bulletin board messages a statement about Lunney because the messages appeared to be by Lunney would set a precedent that allows one person to defame another individual with impunity by putting in that other individual’s mouth words that the first person would be liable for if said directly.[310] Still, harms resulting from misappropriated identity may best be addressed in arenas other than that of libel law. P82. If the Court of Appeals concludes that the Appellate Division is incorrect in finding that the messages are not, as a matter of law, of or concerning Lunney, the case must be examined under the “defamatory,” “publication,” “fault,” and “injury” elements of libel.[311] If none of these elements can be dismissed as a matter of law, the case should go to trial. P83. The messages may not be defamatory as a matter of law if they cannot reasonably be construed as anything more than meaningless vulgarity.[312] As with the “of and concerning” element, however, the common law does not apply rigid requirements of form on the “defamation” element, and necessitates only that someone reasonably construe from the statement a defamatory meaning.[313] The court could conclude that a person (such as the Scout Leader) might reasonably infer from the e-mail messages that Lunney was threatening battery and rape.[314] P84. The publication element seems to be met in light of the adage under traditional libel analysis that “every one who takes part in the publication ... is charged with publication.”[315] Some commentators argue that service providers should not be deemed to have published third-party communications, especially e- mail, but should be placed alongside telephone companies in the third category of participants discussed in Part II.A.1.[316] The ability of a service provider to exert editorial control, if it so chooses, and to restrict or refuse a party’s use of its services, belies that analysis.[317] P85. As for the type of publisher, Prodigy would seem to have been acting as a distributor rather than a primary publisher in light of evidence that it did not exercise significant editorial control over the e-mail and bulletin board messages.[318] Thus Lunney is faced with demonstrating that Prodigy knew or should have known of the e-mail and bulletin board messages’ defamatory content. Note that if Prodigy is deemed a distributor with knowledge of the defamatory content, or a primary publisher, the degree of fault necessary to find Prodigy culpable may be low, as this appears to be a case involving a private-figure plaintiff, a matter not of public concern, and a non-media defendant.[319] Lunney’s claim might nonetheless fail for lack of proof of special damages if the e-mail and bulletin board messages are found not to be libel per se.[320] P86. Analyzing the case this way would help clarify application of the libel-related provisions of the CDA or, alternatively, application of traditional libel analysis to service providers. This approach would also avoid drawing what is at this time an ill-advised analogy between service providers and common carriers. As tempting as it may be to draw such an analogy, courts should refrain from doing so in light of the reluctance of Congress and the Federal Communications Commission to declare service providers to be common carriers, with the concomitant obligation to serve all indiscriminately and without restriction. Unless and until Congress or the FCC does so, the analogy is strained at best, and could have unanticipated consequences. FOOTNOTES [*] Bachelor of Science in Journalism, 1991, Northwestern University, Medill School of Journalism, Evanston, Ill.; Juris Doctor, 1994, Washington University Law School, St. Louis, Mo.; Attorney, Federal Communications Commission, Common Carrier Bureau, Washington, D.C. The views expressed in this article are my own, and should not be attributed to the Federal Communications Commission. I thank Scott Gant for his invaluable support and advice in the writing of this article. I also express my gratitude to J. Robert Lunney of Lunney & Murtagh L.L.C. and Michael J. Silverberg of Phillips Nizer Benjamin Krim & Ballon L.L.P. for providing background on the Lunney case. I am indebted, as well, to John Shaugnessy and Kyle Christensen of the West Publishing Co. for generously providing access to WestLaw. I also thank Karen Sandler and the staff of the Columbia Science and Technology Law Review, and congratulate them on the publication of their inaugural issue. [1] The CDA defines an “interactive computer service” as “any information service, system, or access software provider that provides or enables computer access by multiple users to a computer server, including specifically a service or system that provides access to the Internet and such systems operated or services offered by libraries or educational institutions.” 47 U.S.C. § 230(e)(2) (Supp. III 1997). For further discussion of the CDA, see infra Part II.B.3. [2] In New York, the Supreme Court is both a trial and appellate court. First appeals are heard in one of the Supreme Court’s appellate divisions. The state’s court of last resort is its Court of Appeals. [3] See Lunney v. Prodigy Servs. Co., 683 N.Y.S.2d 557 (N.Y. App. Div. 1998), leave to appeal granted, 93 N.Y.2d 809 (N.Y. argued Oct. 13, 1999). [4] The e-mail message, sent to Lunney’s Boy Scout troop leader, read: “You piece of Shit. I’m Fucking Alex Lunney of troop one and I’m gonna’ kick your ass, you fat piece of shit! I want to butt fuck your sons one at a time and then I’m gonna’ show your wife how a real Boy Scout pitches a tent!” Prodigy Appellate Division Brief at 5. The bulletin board messages, directed at other Prodigy users, read: “Hey Jennifer, blow me!” and “Yeah Mike. She gets home from the whorehouse at noon, you can order her services then.” Id. at 7. [5] The N.Y. Court of Appeals is the state’s highest court. See supra note 2. [6] See 1 ROBERT D. SACK, SACK ON DEFAMATION § 2.3, at 2-7 (3rd ed. 1999); W. PAGE KEETON ET AL., PROSSER AND KEETON ON TORTS § 111, at 771, § 112, at 785 (5th ed. 1984). Injury to reputation caused by oral communication is referred to as slander. 1 SACK, supra, § 2.3, at 2-7; KEETON, supra, § 111, at 771, § 112, at 785. For further discussion of the distinction between libel and slander, see generally KEETON, supra, § 112, at 786-88. “Most commentators agree that defamation on the Internet is best classified as libel.... However, no judicial opinion has squarely addressed whether online speech is libel or slander, and the answer may vary according to jurisdiction.” Developments in the Law, The Law of Cyberspace: III. The Long Arm of Cyber- Reach, 112 HARV. L. REV. 1610, 1612 n.9 (1999) (citations omitted) (citing articles). [7] Because the requirements for a cause of action in libel differ depending upon a variety of factors, at least one author has argued that stating the elements is of limited use other than as a list of issues to address. 1 SACK, supra note 6, § 2.1, at 2-1 to 2-2. [8] Id. § 2.5.1, at 2-72, 2-73; KEETON, supra note 6, § 113, at 797; 43A N.Y. JUR.2D Defamation and Privacy §§ 4, 72 (1994). [9] 1 SACK, supra note 6, § 2.5.1, at 2-73 to 2-74; KEETON, supra note 6, § 113, at 803. [10] See KEETON, supra note 6, § 113, at 799 (stating that “every one who takes part in the publication … is charged with publication”). Individuals can, for example, be deemed to have published the statements of their agents. 1 SACK, supra note 6, § 2.10.2, at 2-136 to 2-138, § 5.5.2.5, at 5-91 to 5-92. This issue has arisen in at least one on-line libel case, Blumenthal v. Drudge. See infra text accompanying note 168. Issues of agency can similarly arise in determining the amount of editorial control a defendant has over a statement, which also is sometimes an issue in on-line libel cases. See infra text accompanying notes 73-78 and 97- 98. [11] See KEETON, supra note 6, § 113 at 803. [12] See, e.g., Tacket v. General Motors Corp., 836 F.2d 1042, 1046-47 (7th Cir. 1987) (concluding that whether General Motors could be deemed to have published a defamatory sign that remained on the wall of one of its plants for between seven and eight months was an issue for the jury because a reasonable person could conclude that General Motors intentionally and unreasonably failed to remove the sign). See also SACK, supra note 6, § 2.5.1, at 2-74 to 2-75 & n.320 (citing Tacket and similar cases, and arguing that but for passage of the Communications Decency Act, such a rationale could have applied to computer bulletin boards); KEETON, supra note 6, at 801. For discussion of the Communications Decency Act, see infra Part II.B.3. [13] See KEETON, supra note 6, § 113, at 803, 810. [14] Id. at 804, 810. [15] Id. at 810. For discussion of fault, see infra Part II.A.7. [16] KEETON, supra note 6, § 113, at 803, 810-11. [17] Id. at 804, 810-11. See also Macaluso v. Mondadori Publ’g Co., 527 F. Supp. 1017, 1019 (E.D.N.Y. 1981). Under this analysis, absence of such knowledge on the part of the distributor would defeat the fault element of the case for libel. At least one case, Church of Scientology v. Minnesota State Med. Ass’n Found., however, suggests that a distributor cannot be deemed to have published the statement absent knowledge of the defamatory content. In Church of Scientology, the court stated that “[t]hose who merely deliver or transmit defamatory material previously published by another will be considered to have published the material only if they knew, or had reason to know, that the material was false and defamatory.” 264 N.W.2d 152, 156 (Minn. 1978) (emphasis added) (citing RESTATEMENT (SECOND) OF TORTS § 581); Hartmann v. American News Co., 171 F.2d 581 (7th Cir. 1948); Balabanoff v. Fossani, 81 N.Y.S.2d 732 (N.Y. Sup. Ct. 1948)). Because such knowledge was lacking, the court concluded that the defendant’s conduct “did not constitute publication.” Id. The Minnesota court seems to have misspoken, as section 581 of the Restatement, as well as the cited aspects of Hartmann and Balabanoff, talk of the absence of fault, not the absence of publication. The U.S. District Court for the Eastern District of Virginia may have fallen victim to similar confusion in Zeran v. America Online Inc. See 958 F. Supp. 1124, 1133 (E.D. Va. 1997) (stating that “distributor liability treats a distributor as a ‘publisher’ of third party statements where that distributor knew or had reason to know that the statements were defamatory”), aff’d, 129 F.3d 327 (4th Cir. 1997), cert. denied, 118 S. Ct. 2341 (1998). For discussion of Zeran, see infra Part II.B.4. [18] See Cubby, Inc. v. CompuServe, Inc., 776 F. Supp. 135, 139-40 (S.D.N.Y. 1991). [19] As the South Dakota Supreme Court put it: It would be ridiculous to say that the author of a libel could escape any liability because malice could not be demonstrated, but that nevertheless, a book seller, or a public library for that matter, could be held liable simply on the basis of knowledge that the publication contains some defamatory material. A book seller is entitled to the same protection afforded by [New York Times v.] Sullivan ....Janklow v. Viking Press, 378 N.W.2d 875, 882 (S.D. 1985). For further discussion of the fault element and the protection afforded by New York Times v. Sullivan, see infra Part II.A.7. [20] KEETON, supra note 6, § 113, at 803-04. [21] Anderson v. New York Tel. Co., 320 N.E.2d 647, 649 (N.Y. 1974) (concurring opinion of Gabrielli, J.). [22] See infra notes 126-132 and accompanying text. [23] Because adoption requires notice of the defamatory content and a reasonable opportunity to act, it fits more squarely into notions of distribution than primary publication. The case for adoption may be considerably harder to make for e-mail messages than the bulletin board messages in light of the fact that e-mail is usually seen only by the sender and the recipients. Courts also may give service providers wide latitude when determining whether their failure to remove e-mail and bulletin board messages is reasonable, considering the difficulty and expense they might experience in doing so and the fact that, depending on the circumstances, readers may be unlikely to infer that the service providers endorse the speech. Cf. Tacket v. General Motors Corp., 836 F.2d 1042, 1046 (7th Cir. 1987) (noting that, because of the “steep discount that readers apply to [defamatory graffiti] and the high cost of hourly repaintings ... [t]he burden of constant vigilance greatly exceeds the benefits to be had” in the context of bathroom walls or subway cars) (citation omitted). [24] KEETON, supra note 6, § 111, at 771. [25] 1 SACK, supra note 6, § 2.5.1, at 2-72 to 2-74; KEETON, supra note 6, § 111, at 771, § 113, at 797-98 & n.1; 43A N.Y. JUR.2D Defamation and Privacy §§ 4, 72, 86 (1994). [26] This element is not really an issue in Lunney’s claims regarding the e-mail and bulletin board messages because they were clearly viewed by third parties. [27] 1 SACK, supra note 6, § 2.9, at 2-114; KEETON, supra note 6, § 111, at 783, § 113, at 802; 43A N.Y. JUR.2D Defamation and Privacy § 4. [28] 1 SACK, supra note 6, § 2.9.1, at 2-115 to 2-116; KEETON, supra note 6, § 111, at 783. [29] KEETON, supra note 6, § 113, at 802; 43A N.Y. JUR.2D Defamation and Privacy § 4. [30] 1 SACK, supra note 6, § 2.4.1, at 2-11 to 2-12; KEETON, supra note 6, § 111, at 774. [31] 1 SACK, supra note 6, § 2.4.7, at 2-36 to 2-37; KEETON, supra note 6, § 111, at 776; Steinhilber v. Alphonse, 501 N.E.2d 550 (N.Y. 1986). [32] KEETON, supra note 6, § 111, at 776, 780, 782. [33] 1 SACK, supra note 6, § 2.4.2.1, at 2-18; KEETON, supra note 6, § 111, at 774, 808-09. But see 1 SACK, supra, § 2.4.3, at 2-24 to 2-25 (stating that the inquiry sometimes rightfully devolves into one of the publisher’s intent). [34] 1 SACK, supra note 6, § 2.1, at 2-3 to 2-4, § 3.1, at 3-1 to 3-2; KEETON, supra note 6, § 116, at 839-42; 43A N.Y. JUR.2D Defamation and Privacy §§ 4, 91 (1994). [35] 1 SACK, supra note 6, § 2.1.1, at 2-4; KEETON, supra note 6, § 116, at 839, 841. [36] 1 SACK, supra note 6, §§ 3.3.1-3.3.2.1, at 3-5 to 3-6, § 3.3.2.2.1, at 3-8 to 3-9. The U.S. Supreme Court has not addressed whether truthful but defamatory statements may be the basis for assessing libel liability regarding statements on matters not of public concern, or made by non-media defendants. Id. § 3.3.2.1, at 3-5 to 3-6, § 3.3.2.1, at 3-8, § 3.3.2.2.2, at 3-9 to 3-11. Thus, states may allow liability for truthful but defamatory statements made by non-media defendants on matters not of public concern without running afoul of the Federal Constitution, although courts rarely if ever find such liability. See id. § 3.3.2.1, at 3-8. The First Amendment requirements act as a floor, and state legislatures and courts may impose falsity requirements in circumstances not mandated by the Federal Constitution. [37] See id. § 1.1, at 1-1 to 1-2; KEETON, supra note 6, § 114, at 815. [38] 1 SACK, supra note 6, § 8.1, at 8-1. [39] Id. § 8.2, at 8-1 to 8-2. [40] See generally id. § 8.2; KEETON, supra note 6, § 114, at 816-24. [41] 1 SACK, supra note 6, § 8.2, at 8-2, § 9.1, at 9- 1; KEETON, supra note 6, § 115, at 825-32. One qualified privilege, the common-interest privilege, applies to communications between people on a subject in which they share an interest, such as the protection of property they jointly own. See generally 1 SACK, supra, § 9.2.3, at 9-20 to 9-25; KEETON, supra, § 115, at 828-30. [42] See 1 SACK, supra note 6, § 9, at 9-4. [43] Id. § 9.1, at 9-1 to 9-2, § 9.3, at 9-29 to 9-30. [44] Id. § 2.2, at 2-5 to 2-6, § 9.3, at 9-29 to 9-30. For further discussion on abuse of qualified privileges, see generally id., § 9.3. [45] Id. § 2.2, at 2-5. [46] Id. § 9.3.1, at 9-32 to 9-33 (emphasis omitted). [47] Id. § 2.2, at 2-6 n.14. For discussion of the term “actual malice” as it is used in the context of determining fault regarding falsity, see infra Part II.A.7. For discussion of the distinction between common-law malice and actual malice, see 1 SACK, supra note 6, § 5.5.1.1, at 5-60 to 5-61, § 9.3.2, at 9-34 to 9-38. [48] See Liberman v. Gelstein, 605 N.E.2d 344, 350 (N.Y. 1992). [49] 1 SACK, supra note 6, § 5.1, at 5-1 to 5-2; KEETON, supra note 6, § 113, at 804, 808-09. [50] KEETON, supra note 6, § 113, at 804. For discussion of truth as an element in libel cases, see supra Part II.A.5. For discussion of privileges, see supra Part II.A.6. [51] 1 SACK, supra note 6, § 2.1.3, at 2-4, § 5.1, at 5-2; KEETON, supra note 6, § 113, at 804-05, 810. [52] 1 SACK, supra note 6, § 5.1, at 5-2, § 5.2.2, at 5-12, § 5.3.1, at 5-14 (citing New York Times v. Sullivan, 376 U.S. 254 (1964); Curtis Publ’g Co. v. Butts, 388 U.S. 130 (1967)); KEETON, supra note 6, § 113, at 804-06, 810. [53] 1 SACK, supra note 6, § 5.1, at 5-2, §6.1, at 6-1 (citing Gertz v. Robert Welch, Inc., 418 U.S. 323 (1974)). Matters of fault arise in libel cases in a variety of ways. KEETON, supra note 6, at 802, 804-08. See also 1 SACK, supra, § 5.1, at 5-3 (observing that fault can be an issue with respect to the “publication,” “of and concerning,” and “defamatory” elements). For example, fault is sometimes an issue in determining whether the plaintiff can be deemed to have intended to publish the statement, as discussed above. See supra text accompanying note 9. Usually when people speak of fault in libel cases, however, they are referring to the defendant’s culpability in not discovering the falsity of the statement. [54] 1 SACK, supra note 6, § 5.3.10, at 5-40 to 5-42, § 6.5, at 6-15 to 6-16 (citing Hutchinson v. Proxmire, 443 U.S. 111 (1979)). [55] Id. § 6.1, at 6-2, § 6.5, at 6-15 to 6-16, § 6.6, at 6-22 to 6-23 (citing Hutchinson, 443 U.S. 111; Dun & Bradstreet Inc. v. Greenmoss Builders Inc., 472 U.S. 749 (1985); Philadelphia Newspapers, Inc. v. Hepps, 475 U.S. 767 (1986)). [56] Id. § 6.4, at 6-11 to 6-12 (quoting Chapadeau v. Utica Observer-Dispatch Inc., 341 N.E.2d 569 (N.Y. 1975)). [57] Id. at 6-15 & n. 101 (citing cases). [58] Id. § 6.4, at 6-14. But see Rupert v. Sellers, 411 N.Y.S.2d 75, 76-77 (N.Y. App. Div. 1978) (4th Dept.) (stating that demonstration of fault is not necessary in cases involving private-figure plaintiffs against non-media defendants on matters not of public concern but concluding, nonetheless, that the evidence in the case at bar could support any finding of fault between negligence and actual malice), aff’d, 408 N.E.2d 671 (N.Y. 1980), cited in 1 SACK, supra note 6, at 6-14 n.99. [59] 1 SACK, supra note 6, § 2.1.4, at 2-5, § 2.4.17, at 2-62 to 2-65, § 2.8-2.8.1, at 2-90 to 2-91, § 2.8.3, at 2-96 to 2-100, § 10.3.3, at 10-8 to 10-10; KEETON, supra note 6, § 112, at 795-97. [60] 1 SACK, supra note 6, §§ 2.8-2.8.1, at 2-90 to 2- 91, § 2.8.3, at 2-96 to 2-97. [61] Id. § 2.8.1, at 2-91. See generally id., §§ 2.8.2- 2.8.3.3. For discussion of slander, see supra note 6. [62] Id. § 2.8.1, at 2-91, § 2.8.3, at 2-96 to 2-98. [63] Id. § 2.8.1, at 2-91, § 2.8.3, at 2-98. [64] See generally id. § 2.8.6.4, at 2-106 to 2-111. [65] Id. at 2-111. [66] Id. § 2.1.4, at 2-5, § 2.8.7.1, at 2-112, § 10.2, at 10-3, § 10.3.1, at 10-5, § 10.3.3, at 10-10; KEETON, supra note 6, § 112, at 796 (citing Gertz v. Robert Welch Inc., 418 U.S. 323 (1974)). [67] 1 SACK, supra note 6, § 6.1, at 6-2, § 10.2, at 10-3, § 10.3.1, at 10-5, § 10.4, at 10-31; KEETON, supra, at 108-09 (5th ed. Supp. 1998) (citing Dun & Bradstreet Inc. v. Greenmoss Builders Inc., 472 U.S. 749 (1985)). [68] See supra Part II.A.7. [69] See Cubby, Inc. v. CompuServe, Inc., 776 F. Supp. 135 (S.D.N.Y. 1991). [70] Id. at 137, 139. [71] Id. at 137, 138. [72] Id. at 137, 138, 139. For a discussion of the difference between distributor and publisher liability, see supra Part II.A.1. [73] Cubby, 776 F. Supp. at 137, 140. [74] Id. [75] Id. at 137. [76] Id. [77] Id. CompuServe users also subscribed directly with Don Fitzpatrick Associates for access to Rumorville, and CompuServe neither paid Don Fitzpatrick Associates for providing Rumorville, nor received any of the fees users paid Don Fitzpatrick Associates for their subscription to Rumorville. Id. [78] Id. at 143. [79] Id. at 140 (citing Daniel v. Dow Jones & Co., 520 N.Y.S.2d 334, 340 (N.Y. Civ. Ct. 1987) (finding computerized database service entitled to same protection as more established means of news distribution)). [80] Id. See also Eric Schlachter, Cyberspace, the Free Market and the Free Marketplace of Ideas, 16 HASTINGS COMM. & ENT. L.J. 87, 134-36 & n.270 (1993) (stating that requiring service providers to continually monitor all the content on their services would be unduly burdensome). [81] Cubby, 776 F. Supp. at 140-41. [82] Id. at 137. [83] Id. at 141. CompuServe subscribers could view Rumorville the instant Don Fitzpatrick Associates uploaded it, providing CompuServe no opportunity to review the newsletter’s contents before it was included in the forum. Id. at 137, 140. [84] Id. at 141. [85] Stratton Oakmont, Inc. v. Prodigy Servs. Co., 1995 WL 323710, at *5 (N.Y. Sup. Ct. May 24, 1995) (citing Edward V. Di Lello, Functional Equivalency and Its Application to Freedom of Speech on Computer Bulletin Boards, 26 COLUM. J.L. & SOC. PROBS. 199, 210-211 (1993)). [86] Id. at *1, *4, *5. [87] Id. at *1. [88] Id. For discussion of libel per se, see supra Part II.A.8. [89] Stratton, 1995 WL 323710, at *1. [90] Id. at *1, *4. [91] Id. at *2, *4. For example, Geoffrey Moore, Prodigy’s Director of Market Programs and Communications, stated in an article in a national newspaper that Prodigy makes “no apology for pursuing a value system that reflects the culture of the millions of American families we aspire to serve. Certainly no responsible newspaper does less when it chooses the type of advertising it publishes, the letters it prints, the degree of nudity and unsupported gossip its editors tolerate.” Id. at *2. [92] For one author’s skepticism regarding the abilities of automatic screening software in general, and Prodigy’s software in particular, see Douglas B. Luftman, Defamation Liability for On-line Services: The Sky is Not Falling, 65 GEO. WASH. L. REV. 1071, 1093-95 & nn.230, 232 & 235 (1997). [93] Stratton, 1995 WL 323710, at *1-2, *4-7. The guidelines asked users not to post “insulting” messages, and informed them that “notes that harass other members or are deemed to be in bad taste or grossly repugnant to community standards, or are deemed harmful to maintaining a harmonious online community, will be removed when brought to PRODIGY’s attention.” Id. at *2. According to a 1993 law review article, Prodigy subscribers were also required to agree to a standard contract in which they promised “not to submit, publish, or display on the Prodigy service any defamatory, inaccurate, abusive, obscene, profane, sexually explicit, threatening, ethnically offensive, or illegal material.” Di Lello, supra note 85, at 206. To enforce the guidelines, Board Leaders could activate an “emergency delete function” to remove a posting and send a number of prepared explanations for the action. Stratton, at *3. Thus, “PRODIGY ha[d] virtually created an editorial staff of Board Leaders who ha[d] the ability to continually monitor incoming transmissions and in fact d[id] spend time censoring notes.” Id. at *5. For further discussion of Prodigy’s censorship practices, see Di Lello, supra, at 207-208; Schlachter, supra note 80, at 94 n.28, 96 & n.34, 102 & n.54, 136 & n.271. [94] Stratton, 1995 WL 323710, at *4, *5. [95] Id. at *3, *4. [96] Id. at *4, *5. [97] See supra notes 74-78 and accompanying text. [98] Stratton, 1995 WL 323710, at *6. For a discussion of Cubby, see supra Part II.B.1. [99] See Luftman, supra note 92, at 1075, 1093, 1095- 96, 1098-99. See also 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, 158-59 (1997) (surmising that the Stratton court was of the opinion that “it was hardly unfair to hold Prodigy to the standard it had set for itself, particularly since Prodigy hoped to increase its profits by setting and publicizing the standard”). Luftman does suggest, however, that the Stratton court’s confusion over the capabilities of the software involved in the case may be partially to blame for the outcome. Luftman, supra, at 1093-95, 1097-98. [100] Stratton, 1995 WL 323710, at *5 (citing Schlachter, supra note 80, at 138-39). Thus, the Stratton court appears a bit inconsistent in its view of content monitoring by service providers. On the one hand, the court criticizes such activity as likely to chill speech, see supra text accompanying note 92, but on the other implies that on-line monitoring has societal benefits that the market may value. [101] Stratton, 1995 WL 323710, at * 5. [102] See Pub. L. No. 104-104, Title V, § 509, 110 Stat. 137 (1996). Other provisions of the CDA that sought to criminalize certain indecent on-line communications were held unconstitutional in ACLU v. Reno, 929 F. Supp. 824 (E.D. Pa. 1996), aff’d, 521 U.S. 844 (1997), and Shea v. Reno, 930 F. Supp. 916 (S.D.N.Y. 1996), aff’d, 521 U.S. 1113 (1997). [103] See 47 U.S.C. § 230(c). According to the legislative history, “[o]ne of the specific purposes of [section 230] is to overrule Stratton-Oakmont v. Prodigy and any other similar decisions ....” H.R. CONF. REP. NO. 104-458, at 194 (2d Sess. 1996) (conference agreement). “[S]uch decisions,” the legislative history continues, “create serious obstacles to the important federal policy of empowering parents to determine the content of communications their children receive through interactive computer services.” Id. [104] 47 U.S.C. § 230(c)(1). The Act defines an “interactive computer service” as “any information service, system, or access software provider that provides or enables computer access by multiple users to a computer server, including specifically a service or system that provides access to the Internet and such systems operated or services offered by libraries or educational institutions.” Id. § 230(e)(2). The Act defines an “information content provider” as “any person or entity that is responsible, in whole or in part, for the creation or development of information provided through the Internet or any other interactive computer service.” Id. § 230(e)(3). The plaintiffs in Blumenthal v. Drudge, discussed in Part II.B.5, infra, latched on to the definition of “information content provider” to argue that section 230 did not protect America Online in that case because they were seeking to hold AOL liable as the content provider and not as “the publisher... of ... information provided by another information content provider.” 47 U.S.C. § 230(c)(1) (emphasis added). [105] 47 U.S.C. § 230(c)(2). Congress amended the act in 1998 to obligate a service provider to notify customers that parental controls such as filtering software are commercially available. See Pub. L. No. 105-277, sec. 1404(a)(2), (3), 112 Stat. 2681 (1998) (codified at 47 U.S.C. § 230(d)). The service provider is to provide this notice when entering a service agreement with a customer. 47 U.S.C. § 230(d). The service provider also must identify, or make available information identifying, providers of such parental control technologies. Id. [106] See 47 U.S.C. § 230(a)(4) (stating congressional finding that “[t]he Internet and other interactive computer services have flourished, to the benefit of all Americans, with a minimum of government regulation”); § 230(b)(1) (stating that it is United States policy “to promote the continued development of the Internet and other interactive computer services and other interactive media”); § 230(b)(2) (stating that it is United States policy “to preserve the vibrant and competitive free market that presently exists for the Internet and other interactive computer services, unfettered by Federal or State regulation”). See also Zeran v. America Online Inc., 129 F.3d 327, 330 (4th Cir. 1997) (stating that “[t]he imposition of tort liability on service providers for the communications of others represented, for Congress, simply another form of intrusive government regulation of speech. Section 230 was enacted, in part, … to keep government interference in the medium to a minimum”), cert. denied, 118 S. Ct. 2341 (1998). For further discussion of Zeran, see infra Part II.B.4. [107] See 47 U.S.C. § 230(a)(2) (stating congressional finding that Internet and other interactive computer services “offer users a great degree of control over the information that they receive, as well as the potential for even greater control in the future as technology develops”); § 230(b)(3) (stating that it is United States policy “to encourage the development of technologies which maximize user control over what information is received by individuals, families, and schools who use the Internet and other interactive computer services”); § 230(b)(4) (stating that it is United States policy “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”). See also H.R. CONF. REP. NO. 104- 458, at 194 (2d Sess. 1996) (conference agreement) (discussing “important federal policy of empowering parents to determine the content of communications their children receive through interactive computer services”); id. (House amendment) (stating intent to “protect[] from civil liability those providers and users of interactive computer services for actions to restrict or to enable restriction of access to objectionable online material”); 141 CONG. REC. S8087- 04 (daily ed. June 9, 1995) (statement of Sen. Exon) (describing goal of earlier version of CDA to limit children’s exposure to indecent material). [108] 47 U.S.C. § 230(e)(3). Section 230(e)(3) also states, however, that “[n]othing in this section shall be construed to prevent any State from enforcing any State law that is consistent with this section.” Id. [109] For discussion of a case suggesting that section 230(c)(1) immunizes service providers even from conduct taken in bad faith, see infra notes 160, and 170-175, and accompanying text. [110] See 1 SACK, supra note 6, § 8.2.7, at 8-41. Despite the opportunity to discuss this question, the U.S. District Court for the Eastern District of Virginia refused in Zeran v. America Online Inc. to address whether section 230 confers absolute immunity. See Zeran v. America Online Inc., 958 F. Supp. 1124, 1133 n.20 (E.D. Va.) (Zeran I), aff’d, 129 F.3d 327 (4th Cir. 1997) (Zeran II), cert. denied, 118 S. Ct. 2341 (1998); Sheridan, supra note 99, at 167, 170-71 (discussing district court decision). For further discussion of Zeran, see infra Part II.B.4. For further discussion of absolute and qualified privileges, see supra Part II.A.6. [111] For discussion of the broad and narrow meanings of “publisher” and the different standards of liability that apply to primary publishers and distributors, see supra Part II.A.1. [112] 129 F.3d 327 (4th Cir. 1997). For a discussion of the different standards of liability applicable to distributors and primary publishers, see supra Part II.A.1. [113] Zeran II, 129 F.3d at 329. [114] Id. [115] Id. [116] Id. [117] Id. [118] Id.; Zeran v. America Online Inc., 958 F. Supp. 1124, 1127-28 (E.D. Va. 1997) (Zeran I). [119] Zeran II, 129 F.3d at 329. Zeran was receiving a threatening call approximately every two minutes by April 30, 1995. Id. Compounding the problem, an Oklahoma City disc jockey received a copy of the first message, and on May 1 recited it on the radio, telling listeners to call “Ken” at Zeran’s number. Id. Because Zeran ran his business from home, he did not want to change his telephone number. Id. The threatening calls got so bad, local police began watching Zeran’s house for his safety. Id. By May 14, 1995, after an Oklahoma City newspaper exposed the hoax and the radio station apologized on air, the calls diminished to fifteen per day. Id. [120] Id. at 328, 329, 330, 334. Zeran had previously filed a separate suit in the U.S. District Court for the Western District of Oklahoma against an Oklahoma City radio station that had repeated the “advertisement” on the air and told listeners to call Zeran. Id. at 329. See supra note 119. Zeran never brought suit against the poster, claiming that AOL’s failure to maintain adequate records of its users made it impossible to identify the culprit. Zeran II, 129 F.3d at 329 & n.1. [121] Zeran II, 129 F.3d at 328, 330. For discussion of the obligation individuals have in certain circumstances to remove material posted on property they control, see supra note 12 and accompanying text. [122] Zeran II, 129 F.3d at 328, 329-30. Even were section 230 not to have barred suit, the court might very well have concluded that the messages were neither “of or concerning” Zeran, nor defamatory, and thus dismissed the case for failure to present essential elements of libel. For discussion of such an argument in Lunney v. Prodigy Servs. Co., see infra notes 221- 223 and accompanying text. For discussion of the “of or concerning” and “defamatory” elements, see supra Parts II.A.3 and II.A.4. [123] Zeran II, 129 F.3d at 328, 330. [124] Id. For the language of section 230(c)(1), see supra text accompanying note 104. Zeran did not dispute the status of AOL as an interactive computer service or the prankster as an information content provider. Zeran II, 129 F.3d at 330 n.2. [125] Zeran II, 129 F.3d at 330. For a discussion of the liability ordinarily applicable to publishers, see supra Part II.A.1. The Zeran court noted that parties can, of course, still seek to hold accountable the original posters of defamatory messages. Zeran II, 129 F.3d at 330. In Blumenthal v. Drudge, the U.S. District Court for the District of Columbia used such a rationale to conclude that rumor reporter Matt Drudge might be held liable for statements in a column of his that AOL carried, but AOL could not be. See Blumenthal v. Drudge, 992 F. Supp. 44 (D.D.C. 1998). For further discussion of the Drudge case, see infra Part II.B.5. [126] Zeran II, 129 F.3d at 331-32. [127] Id. at 331-32. [128] Id. at 332. [129] Id. (citing RESTATEMENT (SECOND) OF TORTS §§ 558(b), 577 (1977); KEETON, supra note 6, § 113, at 799, 802; Tacket v. General Motors Corp., 836 F.2d 1042, 1046-47 (7th Cir. 1987)). [130] Id. (quoting KEETON, supra note 6, § 113, at 799). [131] Id. According to the court, to the extent Stratton and Cubby used “publisher” and “distributor” separately, they appropriately described different liability standards; they did not suggest that distributors are not a subset of publishers. Id. The Zeran court’s conclusion that distributor status is a subset of publisher status for purposes of defamation law made unnecessary any determination of whether AOL was acting as a primary publisher or a distributor. The court continued its analysis, anyway, stating that: The simple fact of notice surely cannot transform one from an original publisher to a distributor in the eyes of the law. To the contrary, once a computer service provider receives notice of a potentially defamatory posting, it is thrust into the role of a traditional publisher. The computer service provider must decide whether to publish, edit, or withdraw the posting. In this respect, Zeran seeks to impose liability on AOL for assuming the role for which § 230 specifically proscribes liability—the publisher role.Our view that Zeran’s complaint treats AOL as a publisher is reinforced because AOL is cast in the same position as the party who originally posted the offensive messages. According to Zeran’s logic, AOL is legally at fault because it communicated to third parties an allegedly defamatory statement. This is precisely the theory under which the original poster of the offensive messages would be found liable. If the original party is considered a publisher of the offensive messages, Zeran certainly cannot attach liability to AOL under the same theory without conceding that AOL too must be treated as a publisher of the statements.Id. at 332-33. This language is somewhat ambiguous. On the one hand, the court may be providing support for its conclusion that distributors are not a distinct category from publishers, but a subset. In this sense, the court would be correct that AOL is acting as a publisher here in the broad sense. On the other hand, the court’s language can be construed as concluding that AOL was acting as a primary publisher, and not as a distributor. This interpretation is difficult to reconcile with traditional libel theory, Cubby, Inc. v. CompuServe, Inc., and Stratton Oakmont, Inc. v. Prodigy Servs. Co. A service provider that offers e-mail, bulletin-board, or database services is a distributor (and thus a publisher in the broad sense) even before it receives notice of an allegedly defamatory statement that has been transmitted over its service; notice simply subjects it to potential liability. See supra note 17 and accompanying text. The service might also be a primary publisher if it exercises editorial control over any of the content it transmits. No evidence is discussed in the Fourth Circuit’s Zeran opinion, however, to suggest that AOL exercised any editorial control over either the specific bulletin board postings at issue or bulletin board postings in general. Thus, the better interpretation would be that AOL was acting as a distributor under traditional libel theory, and that Zeran gave AOL notice of an allegedly defamatory statement. For further discussion of the distinction between primary publishers and distributors, and the broad and narrow senses of “publisher,” see supra Part II.A.1. For further discussion of Cubby, see supra Part II.B.1. For further discussion of Stratton, see supra Part II.B.2. [132] Zeran II, 129 F.3d at 332. Accord, Blumenthal v. Drudge, 992 F. Supp. 44, 52-53 (D.D.C. 1998) (quoting Zeran II, 129 F.3d at 333). See also Doe v. America Online, Inc., 718 So. 2d 385, 388-89 (Fla. Dist. Ct. App. 1998) (quoting Zeran II, 129 F.3d at 330-31) (concluding that section 230 applies to mere distribution, in case seeking to hold AOL liable for allegedly allowing distribution of third party’s child pornography on AOL’s service), review granted, 729 So. 2d 390 (1999). For further discussion of the Drudge case, see infra Part II.B.5. For an argument that Congress meant only to immunize service providers from publisher liability, and not also from distributor liability, see Sheridan, supra note 99, at 167-170. [133] Zeran II, 129 F.3d at 332. [134] Id. at 333, 334. [135] Id. at 333. [136] Id. [137] Id. The Zeran court also attributed to section 230 a congressional intent to prevent tort liability from chilling free speech on-line. Id. at 330-31. According to the Zeran court, notice liability would likely lead service providers to remove potentially libelous material brought to their attention rather than to undertake any on-the-spot defamation and liability analysis. Id. at 333 (citing Philadelphia Newspapers, Inc. v. Hepps, 475 U.S. 767, 777 (1986) (stating that fears of liability can chill speech to the detriment of First Amendment interests)). The court also surmised that notice liability would allow third parties to squelch speech they did not like by claiming it was defamatory, on the theory that service providers would likely remove the content rather than scrutinize all claims. Id. Subjecting service providers to distributor liability might indeed chill speech on the Internet, but the Zeran court’s analysis ascribing to section 230 a congressional motive to protect free speech on the Internet seems somewhat flawed. The court cites congressional findings in the statute that laud the opportunities the Internet provides for diverse discourse, note that it has flourished in a relatively unregulated environment, and express an intent to keep regulation of the internet to a minimum. See Zeran, 129 F.3d at 330 (citing 47 U.S.C. § 230(a)(3), (a)(4), (b)(2)). These provisions stop short, however, of expressing an intent to protect speech. In fact, section 230 arguably makes it easier to squelch speech by immunizing service providers and users from liability for any actions they take in good faith to limit access to material they deem obscene, violent, harassing or “otherwise objectionable.” 47 U.S.C. § 230(c)(2). See also H.R. CONF. REP. NO. 104- 458, at 194 (2d Sess. 1996) (House amendment) (stating intent to “protect[] from civil liability those providers and users of interactive computer services for actions to restrict or to enable restriction of access to objectionable online material”). But see id. (conference agreement) (stating that “[t]he conferees do not intend … that these protections from civil liability apply to so-called ‘cancelbotting,’ in which recipients of a message respond by deleting the message from the computer systems of others without the consent of the originator or without having the right to do so”). [138] Zeran II, 129 F.3d at 328, 334. [139] Zeran I, 958 F. Supp. 1124, 1136. [140] 47 U.S.C. § 230(e)(3). [141] Zeran II, 129 F.3d at 334. [142] Id. at 329, 334-35. [143] Id. at 335 (citing St. Louis v. Texas Worker’s Compensation Comm’n, 65 F.3d 43, 46 (5th Cir. 1995) (stating that a statute that applied to filing of a complaint did not raise retroactivity issue); Vernon v. Cassadaga Valley Cent. Sch. Dist, 49 F.3d 886, 889 (2d Cir. 1995) (same)). See also Doe v. America Online Inc., 718 So. 2d 385, 388 (Fla. Dist. Ct. App. 1998) (finding CDA applicable to complaint filed in January 1997 even though circumstances surrounding complaint occurred in 1994, before enactment of CDA) (citing Zeran II, 129 F.3d at 334, 335), review granted, 729 So. 2d 390 (Fla. April 12, 1999). [144] Zeran II, 129 F.3d at 335 (citing Landgraf v. USI Film Prods., 511 U.S. 244 (1994)). [145] Id. (quoting Landgraf, 511 U.S. at 280). [146] Id. (citing Wright v. Morris, 111 F.3d 414, 418 (6th Cir. 1997) and Abdul-Wadood v. Nathan, 91 F.3d 1023, 1025 (7th Cir. 1996) as cases construing similar statutory language similarly). The appeals court also made it a point to distinguish statutes that impose new liabilities for past conduct from those that affect a litigant’s court access. Zeran II, 129 F.3d at 335. The court observed that judicial opinions “often apply intervening statutes that restrict a court’s jurisdiction. Section 230 neither imposes any new liability on Zeran nor takes away any rights acquired under prior law. No person has a vested right in a nonfinal tort judgment, much less an unfiled tort claim.” Id. (citing Hammond v. United States, 786 F.2d 8, 12 (1st Cir. 1986) (citation omitted)). The appeals court also stated that Zeran took no action in reliance on the status of libel law prior to enactment of section 230. Id. [147] Zeran II, 129 F.3d at 335. It is not clear, however, that Congress had a free-speech interest in mind when it adopted section 230. See supra note 137. [148] See Zeran I, 958 F. Supp. 1124, 1136 & n.27. [149] Id. at 1136 n.27 (emphasis added) (paraphrasing AOL argument and quoting 47 U.S.C. § 230(e)(3)). [150] Id. (paraphrasing AOL and citing Pennsylvania Dept. of Pub. Welfare v. Davenport, 495 U.S. 552, 562 (1990)). [151] Id. [152] Drudge, 992 F. Supp. 44 (D.D.C. 1998). For discussion of the Zeran case, see supra Part II.B.4. [153] Drudge, 992 F. Supp. at 47-48. Drudge would e- mail each new edition to AOL. Id. [154] Id. at 46. Drudge reported that: [n]ew White House recruit Sidney Blumenthal has a spousal abuse past that has been effectively covered up........There are court records of Blumenthal’s violence against his wife, one influential republican, who demanded anonymity, tells the DRUDGE REPORT........One White House source, also requesting anonymity, says the Blumenthal wife-beating allegation is a pure fiction that has been created by Clinton enemies. [The First Lady] would not have brought him in if he had this in his background, assures the wellplaced staffer. This story about Blumenthal has been in circulation for years........Every attempt to reach Blumenthal proved unsuccessful.Id. [155] Id. at 48. [156] Id. at 48 & n.5. [157] Id. at 46. [158] Id. at 49. [159] Id. [160] Id. (quoting Plaintiff’s Memorandum of Points and Authorities in Opposition to Defendant America Online, Inc.’s Motion for Summary Judgment, at 1). See also Sheridan, supra note 99, at 149-50 (observing that, in light of section 230, “a print newspaper could face liability for printing a defamatory letter to the editor, while the publisher of an electronic newspaper would be immune from liability for carrying unedited the same text, even if the publisher of the electronic newspaper acted with the requisite degree of culpability,” and arguing that this result “denies individuals whose reputations have been damaged any recovery against even the largest corporations whose malfeasance or nonfeasance contributed to the damage”). [161] Drudge, 992 F. Supp. at 49-50. [162] Id. at 50. [163] Id. at 51. [164] Id. at 47, 51. AOL had issued a press release captioned “AOL Hires Runaway Gossip Success Matt Drudge.” Id. at 51. The release stated that “[m]averick gossip columnist Matt Drudge has teamed up with America Online.” Id. “Giving the Drudge Report a home on America Online (keyword: Drudge) opens up the floodgates to an audience ripe for Drudge’s brand of reporting,” the release continued. Id. “AOL has made Matt Drudge instantly accessible to members who crave instant gossip and news breaks.” Id. [165] The U.S. District Court for the Eastern District of Virginia seems to have left open the possibility that such a point exists in its consideration of Zeran v. America Online Inc. See 958 F. Supp. 1124, 1133 & n.20 (stating that “there is no occasion here to consider whether, under some set of facts, information initially placed online by a third party might be deemed to be information provided by the service provider itself, thereby rendering § 230(c) inapplicable”). For further discussion of Zeran, see supra Part II.B.4. [166] For further discussion of section 230, see supra Part II.B.3. [167] Drudge, 992 F. Supp. at 50. [168] Id. at 50 & n.9. [169] Id. at 50-51 (quoting Zeran v. America Online Inc., 129 F.3d 327, 330-31 (4th Cir. 1997)). As a mere interactive computer service that carried the Drudge Report, AOL’s role here appears analogous to that of CompuServe in Cubby, Inc. v. CompuServe, Inc. See supra Part II.B.1. Thus, even absent section 230, Cubby might have dictated a similar outcome. The court might have chosen to distinguish the two cases, however, on the grounds that here AOL might reasonably have anticipated the potentially libelous nature of Drudge’s reports. See supra text accompanying notes 81-84. See also supra note 164 and accompanying text. See also Zeran v. America Online, 958 F. Supp. 1124, 1135 (Zeran I) (stating that “distributors of information may be held to have ‘reason to know’ of the defamatory nature of statements made by a third party where that party ‘notoriously persists’ in posting scandalous items”), aff’d, 129 F.3d 327 (4th Cir. 1997) (Zeran II), cert. denied, 118 S. Ct. 2341 (1998). For discussion of Zeran, see supra Part II.B.4. [170] Drudge, 992 F. Supp. at 51. [171] Id. at 51-52 (citing David J. Goldstone, A Funny Thing Happened On The Way To The Cyber Forum: Public vs. Private in Cyberspace Speech, 69 U. COLO. L. REV. 1, 40-48 (1998); Luftman, supra note 92, at 1083-85; Sheridan, supra note 99, at 167-77). See also Schlachter, supra note 80, at 136-38 & n.274, 144 & n.311, 147 (arguing that service providers should be allowed to choose the amount of editorial control they wish to exercise, but should be held to a certain degree of liability depending upon that choice). [172] Drudge, 992 F. Supp. at 52. The court then quoted the provisions of 230(c)(2) that immunize service providers from liability for limiting access to objectionable material, or making it possible for others to restrict such access. Id. For the language of the provisions, see supra text accompanying note 105. The court said that section 230(c)(2) mainly targets obscene and violent material, but stated that the section’s reference to “otherwise objectionable” material was broad enough to encompass defamatory statements as well. Drudge, 992 F. Supp. at 52 n.13. The court’s recitation and expansive reading of section 230(c)(2) appears to have been unnecessary. AOL was accused here of taking insufficient action to restrict access to the communication at issue, rather than exercising the screening or enabling functions that are the subject of section 230(c)(2). See Zeran I, 958 F. Supp. at 1134 n.22 (stating that subparagraph (c)(1) addresses liability for failing to restrict access to material while subparagraph (c)(2) addresses liability for restricting access). The court might have been construing the provisions as conferring on service providers an implicit immunity from liability for decisions not to screen communications or make such screening possible, but reliance on section 230(c)(1) should have been sufficient. A better explanation might be the court’s desire to illustrate the “quid pro quo arrangement” it was describing. This still does not explain, however, the need to interpret the provisions as applying to libel as well as obscene, violent, or harassing communications. What unique purpose section 230(c)(2) would have as applied to libel is also unclear. Section 230(c)(2) by its language would seem only to immunize service providers from liability for actions service providers take to restrict access to objectionable material, which could include some sort of editing function; section 230(c)(1), on the other hand, would seem to apply both to action and inaction that a service provider takes with regard to the content of a third party’s communication. Note also that section 230(c)(2)(A), unlike section 230(c)(1), arguably creates a qualified as opposed to an absolute privilege, and thus can be defeated if the defendant acts with malice. See supra notes 109-110 and accompanying text. [173] Drudge, 992 F. Supp. at 52 (quoting Zeran II, 129 F.3d at 333). For discussion of the Zeran case, see supra Part II.B.4. [174] Id. at 52-53. [175] Id. [176] Id. at 46, 53. [177] The Appellate Division of the N.Y. Supreme Court is a court of first appeal. See supra note 2. [178] Lunney v. Prodigy Servs. Co., 683 N.Y.S.2d 557, 558, 559 (N.Y. App. Div. 1998), leave to appeal granted, 93 N.Y.2d 809 (N.Y. argued Oct. 13, 1999). [179] Lunney, 683 N.Y.S.2d at 558, 559; Prodigy Appellate Division Brief at 5. The message read: “You piece of Shit. I’m Fucking Alex Lunney of troop one and I’m gonna’ kick your ass, you fat piece of shit! I want to butt fuck your sons one at a time and then I’m gonna’ show your wife how a real Boy Scout pitches a tent!” Prodigy Appellate Division Brief at 5. The “from” line listed Lunney’s name and an account number. Id. The subject line read: “How I’m Gonna’ Kill U.” Id. The message was “signed” with Lunney’s name. Id. [180] Lunney, 683 N.Y.S.2d at 558, 559. [181] Id. at 559. [182] Id. [183] Id. [184] Id. [185] Id. It is not clear from the text of the Lunney decision whether the letter was prompted by the e-mail message, or by two offensive electronic bulletin board messages that were posted on Prodigy in Lunney’s name at about the same time. See infra note 193 and accompanying text. [186] Lunney, 683 N.Y.S.2d at 559. [187] Id. At least nine such accounts were revealed during discovery. Lunney Appellate Division Brief at 4. Apparently, anyone stating that he or she was not a minor could open an account by providing a name, address, and—at least since October 1994—a valid credit card number. Lunney, 683 N.Y.S.2d at 559. According to Lunney, Prodigy required no credit card information at the time of the allegedly defamatory posting, had no verification process in place, and could take as long as four months to cancel a fraudulent account. Lunney Appellate Division Brief at 4. [188] Letter from Michael J. Silverberg, Senior Litigation Partner, Phillips Nizer Benjamin Krim & Ballon LLP, to Neil Fried (July 1, 1999) (on file with author); Prodigy Appellate Division Brief at 3. [189] Lunney, 683 N.Y.S.2d at 559; Prodigy Appellate Division Brief at 1, 3, 5. For discussion of libel per se, see supra Part II.A.8. Lunney also argued that Prodigy was culpable for negligence, harassment, and intentional infliction of emotional distress for failing to verify the identity of the individual or individuals that opened the accounts, and for accusing him of sending the offensive messages. See Lunney, 683 N.Y.S.2d at 559; Lunney Appellate Division Brief at 29- 35. For further discussion of those claims, see infra note 221. [190] Prodigy Appellate Division Brief at 3. [191] Lunney Appellate Division Brief at 1. According to Lunney, Prodigy reserved the right to remove or refuse to send messages that violated its guidelines, to inspect electronic messages with the consent of either the writer or the recipient, to do anything it chose with all or parts of public postings, and continues to assert a “‘right to reject any note at any time for any reason.’” Id. at 6. [192] Prodigy Appellate Division Brief at 3. [193] Lunney, 683 N.Y.S.2d at 559; Prodigy Appellate Division Brief at 7. The bulletin board messages read: “Hey Jennifer, blow me!” and “Yeah Mike. She gets home from the whorehouse at noon, you can order her services then.” Prodigy Appellate Division Brief at 7. The “from” lines listed Lunney’s name. Id. [194] Lunney, 683 N.Y.S.2d at 559; Prodigy Appellate Division Brief at 3, 7. Lunney had also learned of, and sought to include in his amended complaint, an electronic bookkeeping entry and a memorandum. Lunney, 683 N.Y.S.2d at 259; Prodigy Appellate Division Brief at 8. The bookkeeping entry was an internal notation Prodigy had made that “[s]ub[scriber] is a non-pay disconnect subscriber 143 days delinquent.” Lunney, 683 N.Y.S.2d at 559; Prodigy Appellate Division Brief at 8 & n.3. The internal memorandum, drafted by a Prodigy subcontractor, stated that “Alex Lunney has been terminated for [credit card] fraud as well as obscene material.” Lunney, 683 N.Y.S.2d at 559. According to Lunney, Prodigy continued to transmit these messages over its network even months after Lunney initiated his lawsuit. Id. [195] Prodigy Appellate Division Brief at 3, 7. For discussion of section 230, see supra Part II.B.3. Prodigy had also argued that the new claims were barred by the statute of limitations. Prodigy Appellate Division Brief at 3, 7. [196] Id. at 3-4. [197] Id. at 4. [198] Id. [199] Id. [200] Id. at 4-5. [201] Id. at 11-16. (citing Cubby, Inc. v. CompuServe, Inc., 776 F. Supp. 135, 141 (S.D.N.Y. 1991)). For further discussion of Cubby, see supra Part II.B.1. [202] Prodigy Appellate Division Brief at 13-16 (citing Anderson v. New York Tel. Co., 320 N.E.2d 647 (N.Y. 1974)). For further discussion of Anderson, see infra text accompanying notes 227-236 and 250-254. [203] Prodigy Appellate Division Brief at 9, 26-28. For further discussion of the requirements that a statement be of and concerning the plaintiff as well as defamatory, see supra Parts II.A.3 and II.A.4. [204] Prodigy Appellate Division Brief at 26. [205] Id. (citing Steinhilber v. Alphonse, 501 N.E.2d 550 (N.Y. 1986)). For discussion of epithets, see supra text accompanying note 31. As for the internal memorandum and bookkeeping entry, see supra note 194, Prodigy argued that they fell within the qualified common-interest privilege. Prodigy Appellate Di