NATO has been using OpenAI’s advanced cyber security models to help protect allied networks, consuming up to 300 billion tokens a month, the company has said.
OpenAI said it works with the NATO Communications and Information Agency (NCIA) to give NATO’s core institutions access to its cyber security models through Daybreak, its programme that puts advanced AI in the hands of trusted defenders for authorised security work. That includes reviewing software, investigating suspicious activity, validating vulnerabilities, and developing and testing fixes.
The announcement came as NATO Edge 26, NCIA’s flagship conference with industry, opened in İzmir.
“A generational leap”
Dr Dylan Browne, NCIA’s General Manager, said: “AI-powered cyber defence models represent a generational leap in cyber security. They accelerate vulnerability discovery, detect emerging threats at machine speed and enable defenders to stay ahead of adversaries.”
“For NATO, these technologies strengthen our cyber resilience, enhance our operational advantage and reinforce our ability to protect critical digital infrastructure at unprecedented scale and speed.”
He said NATO would also study the economics as it scaled up. “As we scale adoption, we will build a deeper understanding of token consumption, usage patterns and the underlying economics of AI models. This insight will help us optimise costs, evaluate competing solutions and make informed procurement decisions.”
OpenAI said NATO’s early use also helped build cyber expertise across the multinational teams protecting allied networks around the clock.
AI agents on Crossed Swords
OpenAI also supported this year’s Crossed Swords exercise, run by the NATO Cooperative Cyber Defence Centre of Excellence in Tallinn. A new training stream saw participants build, deploy and evaluate AI agents for cyber operations using Daybreak, working alongside the agents as operators, with an emphasis on keeping human oversight.
Ukraine and European agencies
Ukraine’s Ministry of Digital Transformation is using Daybreak to strengthen the cyber defence of civilian infrastructure, helping teams find software vulnerabilities and develop and test fixes more quickly.
Speaking at NATO Edge earlier on Tuesday, Ukraine’s digital transformation minister, Oksana Ferchuk, said AI was making Russian cyber attacks “more massive” and “cheaper”, and that defenders “cannot rely on human minds anymore”.
OpenAI said it has also given access to its cyber models to the EU’s cyber agency ENISA and to national cyber agencies in France, Germany and Poland.
ENISA used the models to find four vulnerabilities in software used across EU institutions, with fixes published for all four. A November report from ENISA and CERT-EU reviewed 27 software projects using frontier models, and found that 95% of 88 potential issues assessed by software maintainers were genuine.
Poland’s CERT Polska used OpenAI models to help identify six vulnerabilities in MikroTik’s RouterOS, which is widely used in network routers. MikroTik released fixes, which CERT Polska confirmed stopped the attacks it had observed.









