OpenAI calls for global frontier AI safety standards

OpenAI is calling for global safety standards covering frontier AI systems and is warning against deploying fully autonomous self-improving systems before they can be developed and operated safely.
The proposal reflects growing attention to safeguards for increasingly capable AI models, particularly systems that could independently improve their own capabilities with limited human oversight.
The debate could influence future AI governance and development standards, although the proposed principles do not themselves establish binding regulation.