Lower courts can spend years interpreting the ramifications of US Supreme Court decisions that settle some constitutional matters, leave others unresolved, and never reach arguably adjacent yet distinct statutory questions. That’s happening with the Court’s 2024 ruling in Moody v. NetChoice, as judges examine the relationship among:
(1) the First Amendment expressive rights of social media platforms to curate, organize, and present third-party content as they see fit;
(2) platforms’ use of algorithms to help prioritize, promote, and deliver some of that third-party content; and
(3) the federal statutory protection from civil liability generally afforded to platforms by Section 230(c)(1) for harms caused by publishing third-party content but not for injuries stemming from the platforms’ own speech.
The danger lurking for free-speech advocates, platforms, and most platform users is this: Some judges view Moody as an entrée for eliminating platforms’ Section 230 protection when using personalized algorithms to prioritize and publish third-party content that’s blamed for harm.
Moody. The Court ruled that the First Amendment protects “paradigmatic social-media platforms” against government interference when they create “expressive products” by exercising “editorial judgments” and “mak[ing] choices about what third-party speech to display and how to display it.” Justice Elena Kagan wrote for the majority that Facebook’s editorial decisions about third-party content in its News Feed and YouTube’s homepage feed of third-party videos are protected expressive activities.
The majority never addressed Section 230. As Cato Institute’s Jennifer Huddleston explains, the Court avoided “the question of whether algorithms are protected by an online platform’s Section 230 immunity [and] the full extent of their interaction with First Amendment rights.” The majority only examined whether platforms have First Amendment rights when states compel them to convey content and interfere with “platforms’ choices about whether and how to display user-generated content.” The Court previously “decline[d] to address” whether Section 230 shielded Google from liability for algorithmic-generated recommendations of ISIS videos that allegedly contributed to a terrorist attack.
Fallout. Some judges on the US Court of Appeals for the Ninth Circuit this April in Doe v. Meta Platforms and on the Third Circuit in 2024 in Anderson v. TikTok believe that Moody—a decision safeguarding platforms’ constitutional editorial rights—justifies jettisoning statutory protections afforded under Section 230. As encapsulated in Anderson, the logic is:
Given the Supreme Court’s observations [in Moody] that platforms engage in protected first-party speech under the First Amendment when they curate compilations of others’ content via their expressive algorithms, it follows that doing so amounts to first-party speech under [Section] 230 too.
These judges misguidedly embrace a you-can’t-have-it-both-ways approach to platforms’ legal protections, as if the Court’s recognition that platforms have First Amendment rights when they make “editorial judgments” about “compiling . . . third-party speech” somehow means they can’t also possess statutory immunity as “the publisher” of third-party speech. The constitutional right to curate others’ content into an expressive mosaic, however, doesn’t transform any piece of assembled third-party content into the platform’s own, first-party speech for statutory purposes. These are two different realms of law.
The civil-liability immunity provided by Section 230 for harms caused by publishing third-party speech has solidified over decades through lower-court rulings. Moody, however, provides some judges with a rationale for engaging in what Professor Eric Goldman called in Anderson “a terrible example of judicial activism” by trashing precedent.
Another problem (one telegraphed by concurrences in Doe v. Meta Platforms) rests in how to interpret who’s a “publisher” under Section 230 and what constitutes publishing conduct. Section 230, adopted 30 years ago, affects fast-moving, constantly evolving computer and internet technologies. Platforms, as speech intermediaries, by necessity use algorithms to implement—at massive scale—human judgments about prioritizing and publishing third-party content that users enjoy.
In their Doe concurrences, Judges Marsha Berzon and Ryan Nelson would unravel Section 230’s safeguards partly by turning back the clock. They variously rely on notions of “traditional publishing activities,” “traditional publication,” and an “original public meaning of the term ‘publisher’” that is “consistent with the traditional publisher-distinction developed over centuries” through “the history and tradition of common-law protection for publishers.” While this approach to interpreting Section 230 squares with Justice Clarence Thomas’s logic in a solo, nonbinding 2020 opinion, it misses the mark for interpreting a statute affecting modern online publishing in the artificial intelligence era.
Attorneys at Crowell & Moring accurately assert that Berzon and Nelson’s concurrences “reflect judicial willingness to reconsider a key term in Section 230 that could reshape modern social media law” and “would treat modern algorithms tailored to individual user preferences as falling outside the scope of Section 230 immunity.” They add that the concurrences “foreshadow what may become the most consequential legal debate about the internet in a generation.”
Anderson and Doe dangerously prime this for Supreme Court resolution just as public antipathy, lawsuits, and legislation batter platforms.