Chris Lehane, OpenAI's Chief Global Affairs Officer, expressed the view that AI capabilities have reached a new stage and that AI policies commensurate with this progress are necessary. He emphasized the importance of sharing standards on when development should be slowed, as well as collaborating on technical safety measures, while simultaneously accelerating the development of AI technology.
OpenAI Urges Urgency in AI Safety Policy, Expresses Support for Four California Bills
This article is a translation. Read the Japanese original
OpenAI maintains that ensuring safety is not a hindrance to technical progress, but rather the foundation that enables broad societal implementation. The company stated that it is strengthening monitoring and security measures throughout the model development lifecycle and intends to delay development or deployment if inappropriate behavior occurs during development. Furthermore, regarding "recursive self-improvement," where AI repeatedly improves itself autonomously, the company stated that this should not be pursued unless safety can be ensured.
Regarding regulation, OpenAI proposed the need for mandatory national regulations targeting resource-rich laboratories developing frontier AI. The company expressed support for four bills that have passed the California legislature and are expected to be signed by Governor Gavin Newsom. These include risk assessment processes for AI by independent organizations, the registration and transparency of AI auditors, safety measures for minors using chatbots, and the strengthening of screening standards for biological threats.
Sources
- The AI policy window is open. We need to act. （OpenAI News, 2026-09-09）