Proposed AI Regulation Bill in California Stirs Infighting Among Developers

SB 1047, a bill aimed to protect against the supposed dangers of artificial intelligence, may become law in California. However, the measure has caused some conflict among AI developers.

The proposed Safe and Secure Innovation for Frontier Artificial Intelligence Models Act is a direct aim at developers spending more than $100 million on an AI model. Under the measure, developers must adhere to various requirements, such as safety testing and third-party audits. In addition, the bill gives California’s attorney general the power to take action if an AI model causes significant harm and offers protection for whistleblowers. Model creators must also install code that can turn off the technology at a moment’s notice.

“The godmother of AI,” Dr. Fei-Fei Li fears the proposed law puts unnecessary restrictions on AI developers. If enacted, such legislation could place liability not only on the party misusing the technology but also on the AI coder, which could dampen innovation.

“SB 1047 will unnecessarily penalize developers, stifle our open-source community, and hamstring academic AI research, all while failing to address the very real issues it was authored to solve,” she wrote in Fortune.

In a letter to SB 1047’s co-author California State Senator Scott Wiener, OpenAI’s Chief Strategy Officer Jason Kwon opposed the bill, writing that it would inhibit AI progress and force many companies to move outside California. Kwon says AI policy should be left to the federal government instead of a mixed bag of various state laws.

Former OpenAI researchers Daniel Kokotajlo and William Saunders noted that OpenAI head Sam Altman advocated for AI regulation. Yet now that the legislation is here, OpenAI seemingly opposes it.

Calling himself an “advocate for AI regulation,” xAI owner Elon Musk showed support for the bill. Musk views the technology as a risk to public safety, which requires some sort of regulation.

“This is a tough call and will make some people upset, but, all things considered, I think California should probably pass the SB 1047 AI safety bill,” he posted on X.

Since being approved by the California Senate, SB 1047 has been amended several times, including the removal of a specific AI policing unit called the “Frontier Model Division.” The State Assembly will vote on it later this week, and if passed, it will go back to the Senate for another vote. It is unclear whether Governor Gavin Newsome will sign it into law.