A call for immediate AI regulation as the policy window closes

OpenAI has urged swift action on AI policy, emphasizing the need for urgent, meaningful measures over perfection. The company highlighted the rapid advancement of AI capabilities, stressing that technical solutions alone are insufficient. OpenAI’s Chief Scientist, Jakub Pachocki, warned of the risks posed by recursive self-improvement, calling for extreme caution. The company has strengthened monitoring, alignment, and security safeguards across the model-development lifecycle, including universal monitoring of full trajectories and mandatory alignment-evaluation gates. OpenAI supports a national framework for AI governance, advocating for mandatory, capability-based regulations that evolve with the technology. The company endorsed four additional California bills, including SB 813, AB 1405, SB 1119, and AB 1864, which aim to establish independent assessments, transparency, and safety measures for AI systems. OpenAI also called for international cooperation, noting that AI capabilities transcend borders and require global standards. The company urged Congress to act before adjournment, emphasizing the importance of safeguarding AI development to ensure public trust and lasting American leadership.