Chris Lehane argues that stronger AI capabilities demand stronger safety evidence, shared standards, and durable policy action while the policy window remains open. OpenAI is pushing for mandatory national AI safety requirements and working with Congress on capability-based regulation. Until federal action, the company supports state legislation, including four California bills: SB 813 for independent safety assessments, AB 1405 on AI-auditor standards, SB 1119 for youth protections, and AB 1864 against AI-enabled biological threats.

OpenAI will collaborate with frontier labs to advance voluntary industry-led standards and advocate for compatible global approaches to measuring capabilities, managing risk, preserving human control, and determining when to slow or stop development. Chief Scientist Jakub Pachocki warns that rapid AI rise, including recursive self-improvement potential, requires extreme caution. OpenAI will continue technical alignment and monitoring efforts, building defensive systems and slowing development when necessary, but stresses that individual lab work is insufficient without shared standards.

Advanced AI could accelerate medicine, strengthen infrastructure, expand opportunity, and solve enduring scientific problems, but risks will spread beyond frontier labs as open models approach today's capabilities. Greg Brockman's concept of a "defenders window" highlights a limited period for defenders to strengthen systems before offensive capabilities spread--policymakers face a similar closing window to establish safeguards before AI outpaces governing institutions. Safety must set the pace of progress, enabling broader benefit.

OpenAI has strengthened monitoring, alignment, and security safeguards across the model-development lifecycle, including universal monitoring of full trajectories and mandatory alignment-evaluation gates for systems like Astra, and will slow or halt development when safety risks are unacceptable.