OpenAI public policy agenda

OpenAI’s mission is to ensure that artificial general intelligence (AGI) benefits all of humanity. Our work is guided by five core principles ⁠ that shape how we build AI and how we engage in public policy:

Democratization. We will resist the potential of this technology to consolidate power in the hands of the few.

Empowerment. We believe AI can empower everyone to achieve their goals, learn more, be happier and more fulfilled, and pursue their dream, and that society as a whole will benefit from this.

It also includes safeguards designed to address the risks teens may encounter when interacting with AI systems, including protections against harmful or age-inappropriate content, clear protocols for responding to serious safety situations, and safeguards against manipulative, or deceptive interactions. Crisis-response protocols should include in-service support, referrals to appropriate real-world resources such as 988 in the United States, and timely parental notifications when appropriate. Companies should also be required to protect teen's personal information, including by restricting targeted advertising to minors and the sale of personal data.

The following priorities reflect how we translate our mission ⁠ and our principles ⁠ into public policy. We support policies that help governments and society respond to the opportunities and challenges created by AI, mitigate risks, expand access to opportunity, and ensure people can meaningfully participate in and benefit from the AI economy. Together, these priorities are intended to help governments better serve their constituents while advancing our mission of ensuring that AGI benefits all of humanity. They represent some of the areas where we are most actively engaging today, but this list is not exhaustive. As AI evolves, we also expect our policy priorities and areas of engagement to continue to evolve as well.

We believe frontier AI safety is a national security and public safety issue, particularly for the most advanced general-purpose AI models, which can create risks related to cyber, chemical, biological, radiological, and nuclear (CBRN) weapons. In the US, we support state efforts to align around common frameworks such as California SB 53 ⁠ (opens in a new window) ,the New York RAISE Act ⁠ (opens in a new window) , and Illinois SB 315 ⁠ (opens in a new window) which emphasize transparency, public reporting around catastrophic-risk evaluations and safety incidents, whistleblower protections, and enforceable accountability for developers that fail to meet their safety and security responsibilities. These state-level approaches can help establish harmonized standards that reduce fragmentation and create a path toward an eventual federal framework.

We also support Congressional action to establish a comprehensive federal framework that leverages the emerging consensus reflected in state frontier safety laws; strengthens the Center for AI Standards and Innovation (CAISI) as the US federal government's primary institution for frontier AI safety; and mobilizes a broader resilience plan across government to address the national security and public safety challenges posed by frontier AI. This framework should require CAISI to conduct evaluations of the most capable frontier models, direct CAISI to create an independent assessment ecosystem, and prioritize monitoring progress towards recursive self improvement (RSI). With a comprehensive federal framework in place, we support the preemption of state laws that seek to regulate the same frontier safety risks.

Universal prosperity. We want a future where everyone can have an excellent life.

Resilience. AI will introduce new risks, and we will work with other companies, ecosystems, governments, and society to solve them.

Adaptability. We continue to believe the only way to meet the challenges of a very unpredictable future is to be prepared to update our positions as we learn more.