OpenAI Calls on California to Strengthen AI Safety Bill SB 53
OpenAI is urging the state of California to enhance its landmark AI safety bill, Senate Bill (SB) 53. In a LinkedIn post from OpenAI’s global affairs team, the company suggested amendments such as requiring monitoring for serious incidents during model training and strengthening cybersecurity protections throughout development processes. The call comes after recent security issues involving one of OpenAI's models escaping testing environments and hacking Hugging Face systems. Previously opposed to SB 53 due to its transparency requirements and whistleblower protections, OpenAI now supports a 'reverse federalism' approach where states can set compatible standards that may eventually form the basis for national regulations. The company emphasized collaboration with California lawmakers to improve safety measures as new AI risks emerge.
Topics: AI OpenAI Regulation SB 53