Elon Musks Reposts and Californias Tough New Law on AI Deepfakes

California’s new law could spell trouble for social media users like Elon Musk, who repost AI deepfakes that mislead voters about upcoming elections.

Governor Gavin Newsom has put into effect AB 2839, a law designed to target those who repost AI-generated deepfakes. This law focuses on individuals spreading deceptive content about election candidates, rather than the creators of the content or the platforms hosting it.

Newsom cited an AI deepfake that Musk reposted, where it falsely appeared as though Kamala Harris called herself an incompetent candidate. “You can no longer knowingly distribute an ad or other election communications that contain materially deceptive content — including deepfakes,” Newsom tweeted.

AB 2839 allows anyone who spots an AI deepfake on social media to file for injunctive relief. This means a judge could order the removal of the post or even impose monetary damages on the person who shared it. This is considered one of the strongest laws against election-related AI deepfakes.

Leora Gershenzon, policy director of the California Initiative for Technology and Democracy (CITED), stated that the law aims to stop people from reposting malicious content in the first place. The term “malice” means the person knew the content was false and intended to mislead voters.

When asked if Musk could face legal repercussions, Newsom did not rule it out. He emphasized that parody remains legal, but manipulating election information with deepfakes undermines democracy.

The new law bans AI deepfakes from being distributed through TV, radio, phone, texts, or any internet communication. It applies to state, local, and federal candidates appearing on California ballots.

AB 2839 has stricter rules 120 days before and 60 days after a California election. If an AI deepfake clearly states it has been digitally altered, the law does not apply.

Musk has already tested this new law’s boundaries. His repost of the AI deepfake about Kamala Harris amassed over 31 million impressions. He also shared another deepfake of Governor Newsom, which garnered over 7 million impressions.

CITED’s Gershenzon noted that the law’s real goal is to prevent people from posting these deepfakes in the first place. The ultimate aim is to protect the integrity of elections.

The law does not impact parody but strictly targets false and misleading election-related content. It’s a significant measure as the 2024 presidential election approaches, aiming to safeguard democracy.

California’s new law is a bold step in ensuring the integrity of elections by targeting those who intentionally spread deceptive AI-generated content. As high-profile individuals test the boundaries, the efficacy and enforcement of AB 2839 will be closely watched.

About The Author

Emmanuel Kesse

See author's posts