TLDR:

  • OpenAI opposes California’s AI safety bill SB 1047, claiming it would stifle innovation

  • Senator Scott Wiener defends the bill, saying OpenAI’s arguments are unfounded

  • The bill requires AI companies to conduct safety evaluations before public release

  • OpenAI and other tech companies prefer federal regulation over state laws

  • The bill has been amended but still faces opposition from the tech industry

OpenAI, the artificial intelligence company behind ChatGPT, has officially expressed opposition to California’s Senate Bill 1047 (SB 1047), a proposed law aimed at regulating AI technologies.

The bill, introduced by State Senator Scott Wiener in February 2024, requires AI companies to conduct rigorous safety evaluations of their models before releasing them to the public.

In a letter addressed to Senator Wiener and Governor Gavin Newsom, OpenAI’s Chief Strategy Officer Jason Kwon argued that SB 1047 would threaten California’s economic growth, slow the pace of innovation, and potentially drive AI talent out of the state.

Kwon emphasized that the AI revolution is just beginning and that California’s status as a global leader in AI is fueling the state’s economic dynamism.

OpenAI’s stance aligns with broader pushback against the bill from other tech industry players, including trade groups representing Google and Meta, investment firm Andreessen Horowitz, and prominent AI researchers.

The company joins the chorus of voices calling for federal regulation instead of state-level laws, arguing that a national approach would provide more clarity and certainty for AI developers while preserving public safety.

Senator Wiener, however, quickly responded to OpenAI’s opposition, stating that their arguments “make no sense.”

He pointed out that OpenAI’s letter does not criticize any specific provisions of the bill and that the company’s claim about companies leaving California is unfounded. Wiener emphasized that SB 1047 would apply to all AI model developers doing business in California, regardless of where they are headquartered.

We’ve received valuable feedback from AI developers, startups, and academics on SB 1047. Based on these conversations—& in addition prior big amendments—we’ve made new amendments to help improve the bill as it heads to a full vote in the Assembly. Here are the key changes:🧵 pic.twitter.com/gZLIkBVoYe

— Senator Scott Wiener (@Scott_Wiener) August 16, 2024

The senator acknowledged that while federal regulation would be ideal, Congress has not yet taken action on AI safety.

He drew parallels to California’s data privacy law, which was passed in the absence of federal legislation and has since become a model for other states.

SB 1047 has undergone significant amendments since its introduction. The bill now focuses on requiring AI companies to perform comprehensive safety evaluations on their models to identify potential risks before release. It also grants authority to shut down models that pose significant risks.

These changes were made in an attempt to address some of the tech industry’s concerns and make the bill more palatable.

Despite these amendments, OpenAI and other tech companies remain opposed to the legislation. They argue that the responsibility for misuse of AI tools should rest with those who abuse them, not with the developers who create them.

Some critics, including former House Speaker Nancy Pelosi, have called the bill “well-intentioned but ill-informed.”

The bill has passed through California’s Appropriations Committee and is now headed for a final vote in the state’s Assembly.

If passed, it could land on Governor Newsom’s desk by the end of August 2024. The governor has not yet indicated his position on SB 1047, but signing it into law would likely face significant industry backlash.

Proponents of the bill, including Senator Wiener, maintain that SB 1047 is a necessary step toward responsible AI governance.

They argue that the potential risks associated with advanced AI systems warrant proactive regulation to protect public safety and national security.

The post California’s AI Safety Bill Faces Opposition from OpenAI and Tech Industry appeared first on Blockonomi.