OpenAI has unveiled its "National Security Principles," outlining the permissible use of its AI technologies in collaboration with national governments, particularly in defense and law enforcement sectors. The principles explicitly prohibit the use of OpenAI's technology for mass domestic surveillance, autonomous weapon systems, and high-risk automated decision-making. The company has established "trusted access" partnerships with the U.S. and allied nations, focusing on cybersecurity and biosecurity. This includes collaborations with national cybersecurity agencies through the Daybreak program and providing its biosecurity model, GPT-Rosalind, to select government agencies for public health and biodefense initiatives. OpenAI also advocates for legislative measures to regulate high-risk military applications of AI.