A federal judge blocked one of California's newartificial intelligence laws, just two weeks after it was signed into law by Governor Gavin Newsom. AB 2839, designed to regulate the distribution of AI deepfakes on social media, was blocked following a lawsuit brought by X platform user Christopher Kohls, who had posted a deepfake of Vice President Kamala Harris.
AB 2839 and its Controversy
AB 2839 targets people who distribute deepfakes on social media, especially those that depict political candidates and could mislead voters. Unlike other regulations, AB 2839 does not focus on the platforms where these deepfakes are posted,but on those who share them. The law allows California judges to order the removal of such content or impose fines on violators.
After the law was signed, Newsom suggested thathe could force Elon Musk to remove a deepfake of Kamala Harris that Musk had republished, sparking an online feud between the two. However, Kohls filed a lawsuit the next day, arguing that his post was satire protected by the First Amendment of the US Constitution.
The Court Decision
On Wednesday, federal Judge John Mendez ruled in Kohls' favour, issuing a preliminary injunction thattemporarily blocks the California attorney general's enforcement of the law. Mendez concluded that the law, as written, is overly broad and could unduly limit free speech. Although the order only temporarily blocks the law, it excludes audio messages, which still fall under AB 2839.
This ruling is a blow to California's efforts to control the use of deepfakes in the political context, especially ahead of the upcoming elections. Although AB 2839 is just one of 18 new AI-related laws that Newsom has signed in the last month, its impact is on hold until legal challenges are resolved.
The rulingis also a victory for free speech advocates on platforms like X. In the days following the signing of AB 2839, Elon Musk and his allies posted several deepfakes to test the law, sparking this court challenge.
