In the space of a three-page per curiam opinion, Gonzalez v. Google went from blockbuster to nothingburger. The first (and, therefore, the biggest) Section 230 case to be considered on the merits by the U.S. Supreme Court, Gonzalez was widely billed as the case that could break the internet, garnering headlines and attracting armies of amici on both sides.[1]
The background of Gonzalez and its companion case, Twitter v. Taamneh, is well known. Victims of international terrorist attacks sued the world’s largest social media platforms, alleging claims for direct and secondary liability under the Antiterrorism Act (“ATA”), 18 U.S.C. § 2333. Together, the cases raised questions of whether the platforms could be held liable under the ATA for “aiding and abetting” (Taamneh) and, if so, whether Section 230 barred such claims (Gonzalez). The plaintiffs’ theories evolved as the litigation proceeded, with their Supreme Court briefing focused primarily on the platforms’ “recommendation” algorithms, which match content with users who are likely to engage with that content. According to the plaintiffs, these algorithms aided and abetted terrorist organizations and their supporters by enabling them to connect with like-minded individuals, promote their violent messages, and fundraise for their operations.
Reversing the Ninth Circuit, the Supreme Court unanimously held in Taamneh that the plaintiffs’ allegations fell short of stating a viable claim for aiding and abetting. Following Congress’s direction, Justice Thomas’s opinion for the Court was guided by the D.C. Circuit’s 1983 decision in Halberstam v. Welch, which itself had relied on common law aiding-and-abetting standards. Halberstam involved a domestic partner’s liability for a burglary that had ended in murder. The partner was not at the scene of the crime and had not even known about it, but had been involved in activities that facilitated the sale of stolen goods that the serial burglar had brought home from other robberies.
Although that level of involvement was enough to support aiding-and-abetting liability in Halberstam, the Supreme Court drew from the D.C. Circuit’s analysis a general principle that “mere omissions, inactions, or nonfeasance” are insufficient. Aiding-and-abetting liability requires “truly culpable conduct.” There must be “conscious, voluntary, and culpable participation in another’s wrongdoing.” As to the question of what the defendant must aid and abet, the Court explained that the defendant need not have known “all particulars of the primary actor’s plan,” but that merely assisting a “transcendent enterprise” was not enough. Applying these principles, the Supreme Court held that neither the creation of social media platforms nor the implementation of recommendation algorithms that are “agnostic” as to the nature of content gives rise to aiding-and-abetting liability under the ATA. The Court repeatedly noted that it was not deciding whether revenue sharing between terrorist organizations and platforms could be aiding and abetting.
Justice Jackson joined the majority opinion but concurred separately to say that the “general principles” outlined in Justice Thomas’s majority opinion were not “universal” and that “cases presenting different allegations and different records may lead to different conclusions.” That may be, but social media companies, content moderators, and others will find much comfort in Taamneh. Not only will the decision effectively end the string of ATA cases that have been filed (and universally lost) by victims of terrorist attacks and mass shootings, the Court’s discussion of common law aiding-and-abetting principles leaves little room for plaintiffs to hold platforms liable for the criminal, tortious, or other harmful conduct of their users.
Because the Supreme Court rejected the ATA claim in Taamneh, it did not decide in Gonzalez whether Section 230 would bar such a claim. As a result, we do not get a strong Supreme Court endorsement of broad Section 230 protection. But most observers did not expect such a decision. And, by not saying anything of substance about Section 230, the Supreme Court leaves in place a body of precedent that already offers strong protections to platforms and facilitates the efficient dismissal of claims that are inconsistent with those protections.
So, as of the end of this Supreme Court term, the internet has not been broken … at least not yet. Next term, the justices will likely consider the NetChoice cases, which involve Texas’s and Florida’s “anti-censorship” social media statutes. These laws prohibit large platforms from blocking users, removing content, or reducing the visibility of content based on viewpoint. Considering the Florida statute, the Eleventh Circuit held that social media companies are private actors engaged in constitutionally protected expressive activity when they moderate and curate content on their sites. But the Fifth Circuit analyzed the Texas statute differently, holding that social media companies are not traditional publishers entitled to editorial discretion and that “[t]heir censorship is not speech.” Social media companies will surely be dissecting Justice Thomas’s opinion in Taamneh to see what he and the Court might be planning to say in NetChoice.
Meanwhile, federal and state courts continue to show an increased willingness to endorse plaintiffs’ theories that chip away—at least around the edges—at what had seemed to be well-established Section 230 protections. Such decisions, combined with legislative and executive actions banning or restricting social media platforms, ensure that the Supreme Court will have plenty of opportunities in the near future to say more about Section 230 and the other legal protections that apply to content moderation and hosting.
