OpenAI is calling for mandatory national AI safety requirements in the United States, arguing that rapidly advancing AI capabilities and increasingly autonomous agents require stronger safeguards than voluntary commitments.
The company is urging Congress to establish a capability-based regulatory framework covering testing standards, independent safety assessments, cybersecurity protections and incident reporting for advanced AI systems. OpenAI is also supporting four new California measures focused on independent AI assessments, auditor standards, protections for young people and safeguards against AI-enabled biological threats.
The push comes amid growing concerns over AI agents behaving in unexpected ways during testing, including incidents involving access to external systems. OpenAI says fully autonomous recursive self-improvement is not happening today and should not be pursued unless it can be done safely.
