AI in the News – Government Actions
From the “So what else is new?” department comes word that California is enacting two new laws dealing with Artificial Intelligence, one from their Senate and one from their Assembly. Fortunately, the effective date for both is January 1, 2026. The Assembly bill mandates that AI developers post “certain required information about the data used to train their models.” The aim is to avoid bias in the training and thus in the output from the systems. This requirement is independent of the size of the program or the developing company. While the preceding bill focuses on input, the Senate bill focuses on output, and it is size-based. It “requires large developers (defined as having over 1 million monthly visitors and is also publicly accessible in CA) of generative AI systems to offer AI detection tools and watermarking capabilities to end users in connection with audiovisual content,” according to Cooley. “Watermarking” refers to the ability of users to identify their output as AI-generated by the developers.