OpenAI has partnered with the NATO Communications and Information Agency (NCIA) to provide cybersecurity AI models that detect bugs, probe suspicious activity, and deploy fixes across Allied networks1,2. NATO has used up to 300 billion raw input and output tokens per month through the arrangement.
Access runs through Daybreak, OpenAI's program for vetted defenders doing authorized work such as reviewing code, investigating suspicious activity, and testing fixes. Government agencies and enterprise customers can access advanced models for cybersecurity work under the program.
ENISA, the EU's cyber agency, used OpenAI models to find four vulnerabilities in software used across EU institutions; fixes have been released for all four. ENISA and CERT-EU reviewed 27 software projects using several frontier models, OpenAI's among them, along with tools such as the Codex security plug-in. Maintainers assessed 88 potential issues and confirmed 95% of them as real.
Ukraine's Ministry of Digital Transformation uses Daybreak to defend civilian infrastructure. France's ANSSI and national agencies in Germany and Poland also have access. At NATO's 2026 Crossed Swords exercise, run by the CCDCOE, participants built and tested AI agents for cyber operations in Daybreak, with an emphasis on keeping humans in charge.
ANALYSIS The 300-billion-token monthly volume and the breadth of participating agencies position Daybreak as one of the largest known government deployments of commercial AI for defensive cybersecurity.
The partnership arrives as AI-driven cyber threats are drawing heightened attention from Western security agencies. The FBI issued a FLASH alert on November 13, 2026, warning that attackers are using the ARTEX agentic AI penetration-testing framework to automate cyberattacks against organizations worldwide[1].
ANALYSIS The NATO deployment and the FBI's ARTEX warning together illustrate a widening contest in which both defenders and attackers are scaling AI tooling for cyber operations.
A statement carried in the reporting noted that as adoption scales, the program expects to build a deeper understanding of token consumption, usage patterns, and the underlying economics of AI models.