US: Lawmakers approve AI regulations banning deepfakes

California lawmakers in the United States (US) have given the green light to a series of proposals aimed at regulating artificial intelligence (AI), according to a report by APThe legislation will help combat the threats posed by deepfakes and prevent individual exploitation by AI.

The California Legislature will vote on several AI regulatory proposals this week as it enters the final week of the session. The bills, whose voting deadline is Saturday, will be sent to Gov. Gavin Newsom. The governor has until Sept. 30 to approve or reject them.

Governor Newsom previously raised concerns that excessive regulation of AI could negatively impact local industries. The proposed legislation addresses critical threats posed by AI, including election manipulation and deepfakes depicting child sexual abuse.

The bills call for the removal of misleading AI-generated material 120 days before and 160 days after Election Day. The legislation targets major social media platforms. Additionally, two specific bills would make it illegal to use AI tools to generate media, both images and videos, that depict child sexual abuse.

In the current regulatory landscape, authorities cannot punish individuals who possess or distribute AI-generated child abuse material if there is no evidence that the subjects depicted in the media are real people. One proposal calls for social media companies to deploy AI detection tools to combat misleading content.

The bills include provisions that would require the disclosure of data used by developers to train their AI models for transparency, to understand how these models operate, and to help prevent threats from persisting. Furthermore, algorithmic discrimination is also a central focus of the proposed legislation, along with establishing safety protocols. AI-generated clones of workers, including voice actors and audiobook readers and translators, would also be banned under the legislation.

These measures are included in the context of similar requirements as those set during the Hollywood demonstrationswhich were led by the Screen Actors Guild-American Federation of Television and Radio Artists (SAG-AFTRA) in 2023. Widespread protests against generative AI shut down Hollywood for weeks last year, with writers and artists calling for action against exploitative practices and AI-facilitated replacements, leading to mounting layoffs and falling wages.

If the legislation passes, California could become the first US state to impose a set of regulations on the use of AI.