OpenAI has published its National Security Principles to explain how it approaches government and national security uses of its AI technology. The move is intended to increase transparency about the conditions and limits under which its systems may be used in sensitive settings.
Why publish the principles?
The company says increasingly capable frontier AI systems are being adopted for important government tasks, including national security. That trend creates new obligations for AI labs, governments and civil society to coordinate on acceptable uses. OpenAI argues that democratic societies should be able to use AI to protect people, defend critical infrastructure, deliver public services, and respond to emerging threats — including areas such as cyber defense and biological security where AI can give defenders a meaningful advantage. At the same time, deployment must reinforce democratic accountability, meaningful human judgment and the rule of law, and avoid concentrating power.
How the document was developed
OpenAI describes the principles as the result of a cross-company effort. The process included engagement with David Kris, a leading national security expert who provided independent judgment, internal listening sessions, and contributions from employees across research, safety, policy and government partnerships teams.
Expanding defensive collaborations
The company says it is expanding work with the U.S. government and allied partners in critical defensive areas, especially cyber and biosecurity. In the past month, as part of its Daybreak cyber defense program, OpenAI established Trusted Access for Cyber partnerships with Australia, Canada, Japan, the Republic of Korea, France, Germany, Poland, the Netherlands and EU institutions including ENISA. It also reports a growing and trusted partnership with the UK government around cyber, testing and evaluation.
OpenAI is applying a similar approach to biosecurity: last month it announced expanded trusted access to its GPT‑Rosalind model for select U.S. government and allied partners supporting public health and biodefense missions.
Limits and existing contractual restrictions
The published principles reaffirm contractual restrictions OpenAI announced earlier this year. Those restrictions prohibit use of OpenAI technology for mass domestic surveillance, for directing autonomous weapons systems, and for high‑stakes automated decisions. OpenAI says these limits are consistent with the principles it has now released.
The company's role in public decision‑making
OpenAI states it is important to be clear and transparent about permissible and impermissible uses of its technology. However, the company also argues that many of the most consequential questions about AI in government — especially in national security — should be decided through democratic processes. Companies should inform and support those decisions rather than make them unilaterally. Accordingly, OpenAI supports legislative efforts to establish safeguards around some of the highest‑risk military uses of AI, including applications related to domestic surveillance, autonomous weapon systems, and other uses that directly affect U.S. national security.
Publication and further information
The company has published the full principles on its website to provide guidance on how it will approach current and future national security and law enforcement partnerships and to clarify the norms guiding its expanding defensive cooperation.
