DISMANTLED
Gonzalez v. Google LLC
598 U.S. 617 (2023) · 2023
The recommendation engine kept its immunity, untested.
“We therefore decline to address the application of § 230 to a complaint that appears to state little, if any, plausible claim for relief.”
The Ruling
In a brief per curiam opinion, the Court declined to address whether Section 230 immunity extends to algorithmic recommendations. It vacated and remanded in light of its companion decision in Twitter, Inc. v. Taamneh, 598 U.S. 471 (2023), which held that the platforms' generic content recommendations did not amount to aiding and abetting under the Anti-Terrorism Act. The result: Section 230's application to recommendation algorithms remains, in effect, unchallenged at the Court.
The Personhood Argument Not Made
Gonzalez was the moment the Court could have decoupled the platform's algorithmic speech from its statutory immunity — and chose not to. By disposing of the case on Taamneh's reasoning, the Court left intact a regime in which the algorithm enjoys First Amendment protection (per Moody) and statutory immunity (per § 230) simultaneously. The personhood asymmetry is doubled: the algorithm is constitutionally a speaker and statutorily not a publisher. The family of Nohemi Gonzalez, killed in the 2015 Paris attacks after ISIS recruitment videos were allegedly recommended on YouTube, was left with no actionable defendant for the algorithmic conduct.
The Execution Gap Created
YouTube serves more than 1 billion hours of recommended video per day (YouTube Official Blog, 2017, figure unrevised in subsequent corporate filings). Recommendation systems drive an estimated 70% of watch time on the platform (statement of YouTube CPO Neal Mohan, 2018). Gonzalez left this entire layer of algorithmic curation immune from civil suit, even where it is alleged to materially contribute to lethal harm.
Primary sources & research
Related cases
Part of The Personhood Prism, the companion to The Execution Gap by Thomas William Hornig. See all personhood cases →