Twenty years ago, in AOL v. Zeran, a three-judge panel of the U.S. Court of Appeals for the Fourth Circuit held that 47 U.S.C. §230 immunized defendant AOL from liability for hosting and failing to block a user’s mendacious electronic bulletin board posts about plaintiff, even after AOL received notice of the existence of the offending posts on its servers. This was the first federal appeals court opinion to define the scope of protection under the Communications Decency Act. This reason alone made the decision and opinion significant. But the Zeran opinion was most notable for its conclusion: an online intermediary may not be held liable for third-party user-generated content, even when it knows that the content is unlawful.

This was an exceptional treatment under law. Generally, under longstanding tort principles, publishers are as liable for distributing material that they know to be unlawful as the original author. But, for the panel, the online services could not be treated in the same way; the internet is different. Without immunity, it explained, intermediaries would be exposed to “liability for each message republished by their services.” That kind of exposure “would have an obvious chilling effect.” “Each notification” of objectionable content, the panel elaborated, “would require a careful yet rapid investigation of the circumstances surrounding the posted information, a legal judgment concerning the information’s defamatory character, and an on-the-spot editorial decision whether to risk liability by allowing the continued publication of that information.” Section 230’s purpose, the Fourth Circuit observed, was to preserve the internet as an open forum for expression and commerce. Imposing intermediary liability for the bad acts of third-party users would undermine that purpose.

Twenty years later, we can safely say that the Zeran formulation has helped to foster diversity and abundance of user-generated online content. In this regard, it has given effect to one of Congress’s chief objectives for §230.

But as the Zeran approach has aged, so has its pertinence, for the worse. Many of the search engines, social media and online marketplaces that benefit from the protection today do far more than serve as “publisher[s] or speaker[s] of any information provided by another information content provider.” The largest and most popular application developers do so much more with their users’ content and data. Google, Facebook and Amazon, for example, do not just publish or edit user content. They design the ways by which their users share information; they analyze and algorithmically sort that information; and they repurpose and commercialize the data in ways that are opaque to most consumers. To talk of publisher or notice liability in this context is quaint and inapposite. It is time that courts attend to the ways in which online intermediaries design their services, rather than reflexively apply publisher liability doctrine.

The Zeran panel also understated (if not misstated) the Good Samaritan purpose of the statute. While it recognized the important statutory objective of “self regulation,” it also speculated that intermediaries would be reluctant to “screen material” because doing so “would only lead to notice” which, in turn, would “create a stronger basis for liability.” To be sure, subsequent opinions by federal courts and state courts of last resort pointedly declined to give the Good Samaritan language in §230(c)(2) real effect. But they cited Zeran as authority. As the first to take up the statute, the Zeran panel’s failure to elaborate the “good faith” language in ways that better achieved the self-regulation objectives of the statute was consequential.

Today, in an irony of ironies, under the Good Samaritan safe harbor, intermediaries do not gain anything from policing their users’ content. Many prominent ones do, of course. But many do not. This has been especially troublesome with regard to services that knowingly host or encourage unlawful user content, including and especially those that materially harm vulnerable people and historically subordinated groups. User-generated nonconsensual porn, advertisements that promote the trafficking of minors, and targeted advertisements that violate laws against racial discrimination in housing and employment are the most notorious examples. Legislators have proposed amendments to redress some of these developments. But it is not at all clear they will succeed; the proposals face substantial pushback from the most powerful application developers.

Now is as good a time as any to recognize the Fourth Circuit’s definitive contribution to the development of the internet generally. In light of today’s state of affairs, however, it is plain that the Zeran framework, the prevailing approach to §230 today, needs substantial retooling.

To be fair, as with most judicial opinions, the Zeran opinion was a creature of its time. In the mid- to late-1990s, writers, business leaders, futurists and cyberprofs were breathless and giddy about how the internet would transform the way in which users learn, develop relationships, govern themselves, and transact business. These early evangelists promised that the internet would upturn or unsettle existing laws, geopolitical boundaries, and government bureaucracies. Their excitement expressed itself in ways that sounded every bit like the deregulatory, free-market worldview in vogue at the time. (Compare, for example, President Bill Clinton’s pronouncement in the 1996 State of the Union Address that “the era of big government is over” with the 1994 manifesto, “Cyberspace and the American Dream,” which asserted that governments in the coming era “will be vastly smaller (perhaps by 50 percent or more) than the current one — this is an inevitable implication of the transition from the centralized power structures of the industrial age to the dispersed, decentralized institutions” of the networked world.) Never mind that this worldview has rightfully always been received with suspicion by people and groups that have relied on zealous federal government oversight to protect them from corporate abuses and insidious forms of systemic discrimination. In retrospect, the argument for a hands-off approach to online content arose from a naïve or indulgently ideological conceit that has not born itself out. Just consider that the largest intermediaries are acquiring far reaching services in new markets and making their positions in the economy and public life indispensable. Along the way, they are pushing the limits of competition law and adjudicating in the first instance which kinds of user information to distribute.

Surely, we could be forgiven for believing that, in its failure to elaborate the Good Samaritan safe harbor, the Zeran panel was shortsighted or, worse, taken by cyberlibertarian triumphalism. Many people were.

In fact, the Zeran panel was tasked with a difficult responsibility: promoting free expression on the one hand and encouraging intermediaries’ “self regulation” of objectionable speech on the other. It was to do this through the doctrinal lens of publisher liability doctrine. Zeran argued that, while the statute identifies “publishers,” it says nothing about distributors. This distinction matters, he posited, because Congress presumably meant to keep distributor liability (or notice liability) fully applicable. The panel, however, rejected the argument, explaining that distributor liability “is merely a subset, or a species, of publisher liability, and is therefore also foreclosed by §230.” Plaintiff’s argument failed, it explained, because “[l]iability upon notice would defeat the dual purposes advanced by §230,” creating “an impossible burden” for online services.

This holding ended the panel’s substantive analysis of the scope of immunity under §230, but should not have. Even while its conclusion may have been correct, the holding begs the question about the Good Samaritan language in §230(c)(2), a provision that invites courts to consider services’ responsiveness to block or removal requests. The provision asserts that intermediaries may not be held liable for voluntarily taking steps “in good faith to restrict access to or availability of … objectionable” content. Relying on this protection, the panel could have determined whether, as pled by Zeran, AOL acted in good faith in its handling of the unlawful content. The outcome might have been the same, but the rule would have been narrower and more consistent with both objectives of the statute. But the panel downplayed the significance of the “good faith” “self regulation” safe harbor, privileging the interest in promoting free expression and commerce.

A narrower rule would have been more adaptable to different business models — even those of today. It would have, on the one hand, imposed the obligation on applications to act in good faith when adverting a policy against unlawful content and, on the other hand, protected applications that serve as true publishers or distributors. The problem of true passive intermediaries knowingly hosting unlawful content would remain, but at least then legislators would have a clear understanding about the right fix.

Twenty years after the Fourth Circuit panel announced it, we might expect that Zeran’s influence would wane, particularly in a market that is as dynamic as that for online applications. But the case remains an important authority across jurisdictions, which is a credit to the persuasiveness of Judge Harvie Wilkinson’s opinion for the panel. It remains above all an important authority on the §230 objective to maintain the robust nature of communications online by minimizing government interference. All we need now is a judicial opinion or, likelier, a legislative enactment that better balances that important statutory objective with the original congressional intention to protect users from unlawful and materially harmful communicative acts online. Only then might we achieve a healthy and robust online environment for communication for everyone.

Olivier Sylvain is an associate professor at Fordham University School of Law.


This essay is part of a larger collection about the impact of Zeran v. AOL curated by Eric Goldman and Jeff Kosseff.