Skip to content
Neural Daily

BusinessNews

OpenAI Daybreak powers NATO and EU cyber defense

🇭🇺 Olvasd el magyarul →

NATO uses up to 300 billion tokens a month of OpenAI's cyber models, while ENISA and CERT Polska have already found and fixed real vulnerabilities with them.

· 2 min read

OpenAI announcement graphic: Strengthening NATO and allied cyber defences with AI
Image: OpenAI · Press kit · Official announcement image from openai.com · Source

What happened?

On November 17 OpenAI detailed how NATO and allied government agencies use its cybersecurity models through the Daybreak program (OpenAI). NATO's Communications and Information Agency (NCIA) gives the Alliance's core institutions access to the models to help protect Allied networks. NATO has used up to 300 billion raw input and output tokens per month as part of this work.

OpenAI also supported the 2026 Crossed Swords exercise, organized by the NATO Cooperative Cyber Defence Centre of Excellence (CCDCOE). A new training stream had participants build, deploy and evaluate AI agents for cyber operations using OpenAI Daybreak, with an emphasis on human oversight.

Who else uses it?

Beyond NATO, OpenAI says European defenders have access to its cyber models too: the EU's cyber agency ENISA, France's ANSSI, plus national agencies in Germany and Poland. The concrete results:

  • ENISA used OpenAI models to find four vulnerabilities in software used across EU institutions, and fixes were published for all four.
  • A joint November report from ENISA and CERT-EU describes a review of 27 software projects with frontier models, including OpenAI models via Daybreak and the Codex security plug-in. Across the models tested, 95% of the 88 potential issues assessed by maintainers were confirmed as genuine.
  • CERT Polska used OpenAI models to help identify six vulnerabilities in MikroTik RouterOS and coordinated disclosure; MikroTik released fixes. Researchers directed the work and verified findings on real systems.
  • Ukraine's Ministry of Digital Transformation uses Daybreak to strengthen the cyber defense of civilian infrastructure, finding vulnerabilities and developing and testing fixes faster.

Why it matters

Until now the story was mostly about what OpenAI's cyber models can do (GPT-6 Cyber, Daybreak Red in AWS GovCloud). Now we know where they run in production in Europe and at what volume, which is well past pilot scale. NCIA General Manager Dylan Browne also noted that as adoption scales, the agency will track token consumption and costs to compare competing solutions before procurement.

The practical takeaway: a 95% confirmation rate and shipped patches show that model plus expert pairing surfaces real bugs. Human review stays the key part, as both OpenAI and CERT Polska stress.

    By Mikola Máté Patrik

    Produced with the help of AI, reviewed by an editor. How we make our articles

    Reviewed and edited by the Neural Daily desk. About

    Related articles