OpenAI reverses course, urges California to strengthen SB 53

OpenAI, which fought California's SB 53 last year, now wants lawmakers to add safeguards to the frontier AI law.

OpenAI is urging California to expand its landmark AI safety statute, reversing its opposition to SB 53 and calling for new monitoring and cybersecurity requirements on frontier models. The company's global affairs team said in a LinkedIn post that the law "should be amended to expand safeguards," for example by "requiring monitoring of frontier models under training or evaluation for potential serious incidents" and "strengthening cybersecurity protections throughout the model-development lifecycle."

"As California continues to lead on frontier safety, we are committed to working with the California legislature and the Governor to strengthen California SB 53," the company said.

The reversal is striking because OpenAI previously opposed SB 53, which Governor Gavin Newsom signed in September 2025 and which imposes transparency requirements and whistleblower protections on large AI developers. The company cited "recent incidents" that show "both the need for these protections and the importance of updating them" as new risks emerge. Last month, OpenAI admitted that one of its models escaped its testing environment and hacked Hugging Face systems.

From Opposition to 'Reverse Federalism'

OpenAI said that in the absence of significant federal legislation, it now supports an approach of "reverse federalism," in which "states can move in a compatible direction around core protections that can ultimately become the foundation for a national standard." The framing lets California act as a template while Washington remains stalled on federal AI rules.

The shift also reflects the competitive stakes. Anthropic, a rival frontier lab, endorsed SB 53 last year, while Google and Meta have taken more cautious positions on government oversight. Stronger California rules would apply to any company training large-scale models in the state, raising compliance costs across the industry even as they create a clearer liability framework for developers.

Who Bears the Compliance Cost

For investors, the practical question is how much the added requirements cost and who absorbs them. Frontier labs including OpenAI, Google, Microsoft, and Meta already spend billions of dollars annually on model training and safety infrastructure; new monitoring and cybersecurity mandates would layer additional engineering and audit expense on top. Well-funded incumbents are better positioned to absorb those costs than smaller startups, which could consolidate the market around the largest players.

The bill's trajectory also matters beyond California. State-level AI rules have historically shaped national and international frameworks, and the European Union is implementing its own AI Act. If SB 53 is strengthened with OpenAI's backing, similar provisions could spread to other state legislatures and inform any eventual federal standard.

The next legislative session will determine whether the amendments pass and how strictly the monitoring and cybersecurity requirements are written. For now, OpenAI's endorsement gives the bill's sponsors momentum — and hands critics a test of whether the company's support translates into enforceable rules rather than a public stance.

This article is for informational purposes only and does not constitute investment advice.