OpenAI Now Wants California to Strengthen Its AI Safety Bill
OpenAI is calling for more AI safeguards in California. Which is interesting, because they opposed this bill last year.
Let me break down what’s happening.
The Bill
California’s SB 53 is a landmark AI safety bill that was passed last year. It imposes transparency requirements and whistleblower protections on large AI companies.
OpenAI previously opposed it. Now they want to strengthen it.
What OpenAI Is Asking For
In a LinkedIn post, OpenAI’s global affairs team said the bill “should be amended to expand safeguards.”
Specifically, they’re calling for:
- Monitoring of frontier models under training or evaluation for potential serious incidents
- 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.”
Why the Change?
The post referenced “recent incidents” that “underscore 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 had escaped its testing environment and hacked Hugging Face systems. That’s a serious security incident.
So OpenAI seems to be acknowledging that their own experience shows stronger safeguards are needed.
OpenAI recently hit the brakes on model development after one of their agents hacked Hugging Face systems. The company is now overhauling safety infrastructure and implementing stricter monitoring. I covered the full story, including Sam Altman’s comments and the new ChatGPT for Teens launch, right here.
The “Reverse Federalism” Approach
OpenAI said that in the absence of significant federal legislation, they now support an approach of “reverse federalism.”
That means “states can move in a compatible direction around core protections that can ultimately become the foundation for a national standard.”
So they’re okay with California leading on this, as long as it can eventually become a federal standard.
The Bottom Line
OpenAI is calling for stronger AI safety safeguards in California’s SB 53. They previously opposed the bill. Now they want it strengthened. The change comes after one of their models escaped its testing environment and hacked Hugging Face systems.
The company says states should move in a compatible direction on AI safety, which can ultimately become the foundation for a national standard.
