California has introduced a series of new laws aimed at cracking down on AI deepfakes in elections and entertainment. On Tuesday, Gov.
Social media companies are struggling to identify and remove harmful, AI-generated deepfakes.
In California, we just enacted new laws to protect both people and our democracy's defenses — and as technology evolves, we won't be slowing down. pic.twitter.com/XKt3l6BlkY
— California Governor (@CAgovernor) September 18, 2024
Gavin Newsom signed five AI-related bills, introducing stringent measures for big online platforms and limiting how studios can exploit the likenesses and voices of performers.
California is standing up to harmful deepfakes in political ads and other online content.
To help safeguard the integrity of our elections, these three new laws will remove deceptive content from large online platforms, increase accountability, and better inform voters. pic.twitter.com/6wUzHpOSGV
— California Governor (@CAgovernor) September 17, 2024
The three bills addressing elections build on existing legislation signed five years ago, which made it illegal to maliciously distribute deceptive media intended to discredit a candidate just before an election. One of the new bills extends this protection from 60 days to 120 days before an election. Another new bill, the Defending Democracy from Deepfake Deception Act, mandates large online platforms to prevent users from posting “materially deceptive” election-related content that falsely depicts a candidate or public official.
"Gov. Gavin Newsom signed legislation Tuesday that would protect actors from being replaced by digital imposters — even after their death — without their consent"https://t.co/2kOE0eTMoJ
— Governor Newsom Press Office (@GovPressOffice) September 17, 2024
Assemblymember Marc Berman, who proposed the bill, said, “Advances in AI make it easy to generate hyper-realistic yet completely fake content, but [the new law] will ensure that online platforms minimize their impact.”
The third bill requires electoral ads featuring AI-generated or “substantially altered” content to include a disclosure.
Looks like SB 1047 will be veto’d by Newsome. Much needed good news for open source AI. 🤞🏼https://t.co/cwlybPXmr6
— JJ (@JosephJacks_) September 18, 2024
This year’s election has already seen misleading AI-generated content, including deepfakes distributed by former President Donald Trump, highlighting the urgent need for these new laws. California’s new laws are particularly notable given the state’s large population and its role as home to major online companies like Meta.
newsom signs AI deepfake legislation
The laws have even sparked reactions from tech moguls like Elon Musk, who criticized them. In the realm of entertainment, California’s new laws are significant victories for SAG-AFTRA, the media professionals’ union.
One law ensures that performers’ voices or likenesses can’t be replicated by AI without their permission, requiring these terms to be included in contracts. Another law protects the digital replicas of deceased performers, ensuring their estates must give consent for commercial use. SAG-AFTRA president Fran Drescher said, “It is a momentous day for SAG-AFTRA members and everyone else.” The new law would allow California’s iconic entertainment industry to “continue thriving while strengthening protections for workers.”
Among AI-related bills awaiting signing is a highly debated one that would force AI companies to ensure their models can’t be used for harmful purposes like biological attacks or major crimes.
The bill has sparked intense debate within the AI community, with figures like OpenAI’s “Godmother of AI” Fei-Fei Li supporting it, while others like Elon Musk and Anthropic have raised concerns. In a press conference, the complexities were acknowledged, stating that the bill could have an “outsized impact” and potentially a “chilling effect” on the open-source AI community. The importance of finding a balanced approach to tackling both demonstrable and potential risks was emphasized.
California’s new deepfake laws mark a significant step in regulating AI-generated content in elections and entertainment. While substantial progress has been made, the debate over the state’s most comprehensive AI bill continues, reflecting the complex nature of AI regulation.
