California legislature bans 'deepfakes,' requires social media platforms to block them around elections

The California legislature concluded its final week in session on Saturday and has approved a number of bills aimed at banning deepfakes and regulating AI in the state. 

One of the laws passed as the deadline loomed is the "Defending Democracy from Deepfake Deception Act of 2024." It requires that a "large online platform," defined as a "public-facing internet website, web application, or digital application, including a social media platform," block "the posting of materially deceptive content related to elections in California, during specified periods before and after an election."

The period set forth by the bill is between 120 days before an election in the state and 60 days after the election. Campaigns will also be required to disclose if they are running ads with materials that have been altered by AI. The bill passed the Assembly on August 28 and the Senate on August 27.

Content covered under the bill includes that believed to be "deceptive content" showing a candidate for office being portrayed as "saying something that the candidate did not do or say and that is reasonably likely to harm the reputation or electoral prospects of a candidate," an elections official "doing or saying something in connection with the performance of their elections-related duties that the elections official did not do or say and that is reasonably likely to falsely undermine confidence in the outcome of one or more election contests," and "An elected official portrayed as doing or saying something that influences an election in California that the elected official did not do or say and that is reasonably likely to falsely undermine confidence in the outcome of one or more election contests."

Another bill passed ahead of the end of session is SB 942, or the California AI Transparency Act. The bill requires that a "covered provider," a person "that creates, codes, or otherwise produces a generative artificial intelligence system that has over 1,000,000 monthly visitors," make available a "AI detection tool" that allows a user to assess when content was "created or altered by the covered provider’s GenAI system."

A covered provider will also be required to include disclosures in AI-generated content stating that the content was AI-generated. The bill passed the Assembly on August 28 and the Senate on August 29.

A second bill related to artificial intelligence passed in the days leading up to the end of the session, the "Safe and Secure Innovation for Frontier Artificial Intelligence Models Act," or SB 1047, sets in place guardrails for training AI. The bill also passed the Assembly on August 28 and the Senate on August 29.

Among the guardrails set in the bill is the requirement to "implement the capability to promptly enact a full shutdown." A developer is also required to implement a written safety and security protocol that specifies the protections and procedures that the developer has taken to "avoid producing a covered model or covered model derivative that poses an unreasonable risk of causing or materially enabling a critical harm," and identifies a testing protocol that "takes reasonable care" to evaluate if the program "poses an unreasonable risk of causing or enabling a critical harm" or its "derivatives pose an unreasonable risk of causing or enabling a critical harm."

California Governor Gavin Newsom has until September 30 to sign or veto the laws.