A federal judge has ruled that California’s “deepfake” election ad ban is unconstitutional, stating that the legislation goes beyond reasonable regulation and infringes on free speech. The decision, authored by U.S. District Judge John Mendez, highlighted multiple constitutional flaws: it suppresses too much speech, imposes content and viewpoint discrimination, and compels expressive content in a selective manner, all of which breach the First Amendment.
California enacted two bills aimed at combating AI-generated political misinformation—AB 2655 and AB 2839—following a satirical deepfake video mocking Vice President Kamala Harris. Governor Gavin Newsom signed the measures, citing concerns over election integrity. Judge Mendez rejected AB 2655 on the grounds that it violated Section 230 of the Communications Decency Act, which protects online platforms from liability over third-party content.
In a broader ruling, the judge also invalidated AB 2839, which permitted any individual to sue over AI-generated political content perceived as “materially deceptive.” Mendez found this to be a blunt, censorship-style approach that could stifle satire and parody. He noted the law only penalized content that could harm a candidate, not content that might benefit one—clear evidence of impermissible viewpoint discrimination.
The court also took issue with enforcement provisions that allowed anyone to file a lawsuit against content creators and mandated overly burdensome disclaimers. Such requirements, Mendez wrote, could effectively kill the joke in a satirical video: “a mandatory disclaimer for parody or satire would kill the joke.”
