California Revises AI Safety Bill SB 1047 Ahead of Final Vote
The Importance of AI Safety Legislation
As AI technology continues to advance at an unprecedented rate, concerns over its safety and potential misuse have grown exponentially. In California, Senator Scott Wiener’s bill, SB 1047, aims to address these worries by holding developers accountable for the dangers posed by their AI models. However, the bill has faced strong opposition from various Silicon Valley stakeholders, including AI firm Anthropic.
Revised Amendments Reflect Industry Feedback
In a significant development, SB 1047 has been revised following input from key industry players, including Anthropic. The amendments, which were proposed in response to concerns raised by the company and others in the industry, aim to address some of the core issues with the original bill. Senator Wiener confirmed these adjustments in a statement to TechCrunch.
Key Changes to SB 1047
The revised bill incorporates several notable changes, including:
- Reduced authority for California’s Attorney General: The provision that allowed the Attorney General to sue AI companies for negligent safety practices before a disaster occurs has been removed. This change means that legal action can only be taken after an incident has occurred.
- Revised oversight and certification requirements: The bill no longer mandates the creation of a new agency, the Frontier Model Division (FMD). Instead, the responsibilities initially assigned to the FMD will be handled by an expanded Board of Frontier Models within the existing Government Operations Agency.
- Softened language regarding safety certifications: Labs are no longer required to submit safety test results under penalty of perjury. Instead, they will provide public statements outlining their safety practices without facing criminal liability.
Additional Amendments Address Open-Source Developers’ Concerns
The revised bill also introduces more lenient language regarding developers’ responsibility to ensure AI models’ safety. The bill now requires ‘reasonable care’ rather than ‘reasonable assurance’ that AI models do not pose significant risks of causing a catastrophe.
A specific protection for open-source developers has been added, which states that developers who spend less than $10 million fine-tuning a covered model are not considered liable under SB 1047. Liability remains with the original developer.
Next Steps and Ongoing Controversy
Despite strong opposition from U.S. congressmen, AI researchers, and major tech companies, SB 1047 has progressed through California’s legislative process with relative ease. The latest amendments are likely to appease some critics, potentially offering Governor Gavin Newsom a less contentious bill to sign.
However, the bill’s core premise—holding developers accountable for the dangers posed by their AI models—remains controversial. Critics argue that the recent amendments do little to address the fundamental issues with the bill.
The Stakes: California’s Startup Ecosystem and Scientific Development
As SB 1047 heads to California’s Assembly floor for a final vote, eight U.S. Congress members from California have urged Governor Newsom to veto the bill, citing concerns over its impact on the state’s startup ecosystem, scientific development, and efforts to protect against potential AI-related harms.
The Future of AI Safety Legislation
If SB 1047 passes, it will return to the Senate for approval of the amendments before heading to Governor Newsom’s desk for a final decision. The outcome will have significant implications for California’s tech industry and the future of AI safety legislation in the state.
Conclusion
The revisions to SB 1047 reflect the ongoing dialogue between lawmakers and industry stakeholders regarding the need for effective AI safety legislation. While some critics argue that the amendments do little to address the fundamental issues with the bill, others see them as a step in the right direction towards advancing AI safety while allowing for innovation across the ecosystem.
As the bill approaches its final vote, it remains to be seen whether Governor Newsom will sign or veto SB 1047. One thing is certain: the future of AI safety legislation in California hangs in the balance.
