Elon Musk has publicly endorsed California Senate Bill 1047, a controversial piece of legislation aimed at introducing safety measures for advanced AI systems. Musk's endorsement, announced via his social media platform X on Monday, comes at a time when the bill has sparked significant debate among tech giants, AI developers, and policymakers.
"This is a tough call and will make some people upset," Musk acknowledged in his post. "But, all things considered, I think California should probably pass the SB 1047 AI safety bill." The Tesla and SpaceX CEO, who also heads the AI company xAI, reiterated his long-standing stance on the need for AI regulation, drawing parallels to the regulation of other potentially hazardous technologies. "For over 20 years, I have been an advocate for AI regulation, just as we regulate any product/technology that is a potential risk to the public," Musk said.
Senate Bill 1047, introduced by California lawmakers, aims to impose stringent safety requirements on AI developers. The bill mandates that developers of large-scale AI models, particularly those with a development budget exceeding $100 million, conduct thorough safety testing before public release. Additionally, it holds developers accountable for any severe harm their models might cause, including mass casualties or damages exceeding $500 million. The bill also includes provisions for a full shutdown protocol for AI models and whistleblower protections.
The proposed legislation has faced strong opposition from major technology companies and AI firms. Giants such as Meta and Google, along with prominent AI developers like OpenAI, argue that the bill's stringent requirements could stifle innovation and hinder the growth of the AI industry in California, a state known for its technological advancements and startup culture.
Former Speaker Nancy Pelosi and other California Democrats have also expressed concerns, sending a letter to Governor Gavin Newsom urging him to veto the bill should it reach his desk. Pelosi and her colleagues worry that the bill could impose excessive burdens on AI developers and potentially slow down technological progress.
Despite these objections, SB 1047 has garnered support from notable figures in the AI research community. Yoshua Bengio and Geoffrey Hinton, two leading AI researchers, have endorsed the bill, praising its focus on safety and accountability. AI startup Anthropic, which had previously criticized the bill, has also shown support following recent amendments that improved the legislation's balance between regulation and innovation. "We believe its benefits likely outweigh its costs," said an Anthropic spokesperson.
The legislation is part of a broader legislative effort in California, where state lawmakers have proposed a total of 65 bills related to AI this session. However, many of these proposals have already failed to advance, making SB 1047 one of the few significant AI-related bills still under consideration.
As the 2024 U.S. presidential election approaches, the potential for AI-driven misinformation and its impact on democratic processes has become a growing concern. Experts have warned about the risks associated with AI's ability to generate and spread false information, underscoring the need for regulatory measures to ensure the technology is used responsibly.
