OpenAI has published its National Security Principles to provide transparency regarding its approach to government and national security partnerships. This initiative aims to ensure that AI deployment reinforces democratic accountability, human judgment, and the rule of law while supporting critical defensive areas like cyber defense and biosecurity.
- The principles were developed through a cross-company effort facilitated by national security expert David Kris, involving listening sessions with employees from research, safety, policy, and government partnerships.
- OpenAI has established Trusted Access for Cyber partnerships with Australia, Canada, Japan, South Korea, France, Germany, Poland, the Netherlands, and EU institutions like ENISA under its Daybreak program.
- The company announced expanded trusted access to its GPT-Rosalind model for select U.S. government and allied partners supporting public health and biodefense missions.
- Existing contractual restrictions with the Department of War prohibit using OpenAI technology for mass domestic surveillance, directing autonomous weapons systems, or high-stakes automated decisions.
OpenAI believes that while companies should help inform decisions, consequential questions about AI in national security must be answered through the democratic process. The company supports legislative efforts to establish safeguards around high-risk military uses of AI.