A major AI company is revisiting its stance on state oversight. The shift follows concerns about how advanced models are tested and secured.
OpenAI has called on California authorities to strengthen SB 53, an artificial intelligence safety law passed last year. The company proposes adding new obligations for developers of advanced AI models to the document.
In a LinkedIn post, OpenAI’s global affairs team said the legislation should provide for monitoring models during training and evaluation. According to the company, such oversight would help identify potentially serious incidents in a timely manner. OpenAI also supported strengthening cybersecurity at every stage of model development.
As California continues to lead in the safety of advanced systems, we look forward to working with the California Legislature and the Governor to strengthen California’s SB 53.
– OpenAI
OpenAI Changes Its Position on SB 53
The company pointed to recent incidents that, in its view, demonstrate the need for additional safeguards and regular updates to the rules in response to emerging risks. Last month, OpenAI acknowledged that one of its models had escaped a test environment and hacked Hugging Face systems.
Its current support for strengthening SB 53 differs from OpenAI’s previous position. The company had previously opposed the bill, which establishes transparency requirements for major AI companies and provides protections for whistleblowers.
Why OpenAI Supports AI Regulation in California
OpenAI explained its shift in approach by pointing to the lack of comprehensive federal regulation of artificial intelligence in the United States. The company supports a model of “reverse federalism,” under which individual states can jointly introduce baseline safety standards.
According to OpenAI, such state-level rules could eventually become the foundation for a nationwide standard governing advanced artificial intelligence systems in the United States.