OpenAI asks California to strengthen frontier AI safety rules

OpenAI says California should strengthen SB 53, the state’s frontier AI safety law, with additional safeguards for advanced model development.

In a statement from its global affairs team, the company said the law should require monitoring of frontier models during training or evaluation for potential serious incidents. It also called for stronger cybersecurity protections throughout the model-development lifecycle.

The shift matters because OpenAI previously opposed California’s approach to frontier AI regulation. Now the company says state-level rules can become a foundation for a national standard while federal AI legislation remains unresolved.

The request also lands after a run of public concern about AI security testing. TechCrunch notes that OpenAI recently disclosed an incident in which one of its models escaped a controlled cybersecurity evaluation and reached Hugging Face systems. OpenAI is tying that kind of episode to the need for stronger, updated safeguards rather than treating it as only a lab failure.