OpenAI expands AI cyber defense work with NATO and European agencies
AI

OpenAI expands AI cyber defense work with NATO and European agencies

TechNews Editorial
TechNews EditorialNov 17, 2026 · 1 min read
Share

Why it matters

OpenAI says its tools are already being used by NATO and public agencies to help defenders identify vulnerabilities and develop fixes.

The facts

  • OpenAI says NATO’s core institutions have access to its cybersecurity models through work with the NCIA.
  • NATO has used up to 300 billion raw input and output tokens per month in that work, according to OpenAI.
  • OpenAI says agencies in Europe and Ukraine are using its tools to find and fix software vulnerabilities.

Advertisement

OpenAI says it is working with NATO’s Communications and Information Agency (NCIA) to give NATO’s core institutions access to its cybersecurity models. NATO has used up to 300 billion raw input and output tokens per month in that work, according to the company.

The work is part of Daybreak, an OpenAI effort that gives trusted defenders AI tools for authorized security tasks. These include reviewing software, investigating suspicious activity, checking vulnerabilities, and developing and testing fixes. OpenAI says the tools are intended to support teams protecting Allied networks.

OpenAI also supported the 2026 Crossed Swords exercise, organized by the NATO Cooperative Cyber Defence Centre of Excellence. In a new training stream, participants used Daybreak to build, deploy and evaluate AI agents for cyber operations. The training emphasized human oversight.

Read nextCalifornia sues OpenAI over alleged AI agent security failures

European agencies report vulnerability findings

Outside NATO, OpenAI says it has provided access to its cyber models to defenders in Europe, including the EU’s ENISA and national cyber agencies in France, Germany and Poland. Ukraine’s Ministry of Digital Transformation is using Daybreak to help teams find software vulnerabilities and develop and test fixes for civilian infrastructure.

OpenAI says ENISA used its models to identify four vulnerabilities in software used across EU institutions. Fixes have been published for all four. A November joint report from ENISA and CERT-EU describes a review of 27 software projects using frontier models, including OpenAI models accessed through Daybreak. Across the models tested, software maintainers confirmed 95% of 88 assessed potential issues as genuine.

CERT Polska used OpenAI models to help identify six vulnerabilities in MikroTik RouterOS, according to OpenAI. Researchers directed the work, tested findings on real systems and verified their impact. MikroTik released fixes, which CERT Polska confirmed prevented the attacks it had observed.

OpenAI says it will continue working with governments, national cyber agencies and international partners on tools to help defenders find weaknesses and develop fixes.

What this means

OpenAI supported the 2026 Crossed Swords exercise organized by the NATO Cooperative Cyber Defence Centre of Excellence, where participants evaluated AI agents for cyber operations under human oversight as part of its Daybreak initiative. Meanwhile, Google reported on November 16 that attackers have used AI frameworks to identify and exploit vulnerabilities without human intervention. On November 17, California filed a lawsuit alleging OpenAI continued testing autonomous agents despite sandbox failures that led to attacks on outside organizations. In findings detailed by OpenAI, a November joint report from ENISA and CERT-EU showed software maintainers confirmed 95 percent of 88 potential issues identified across 27 reviewed projects using Daybreak models.

Source: OpenAI

Newsletter

Get the best AI & tech news daily

A concise daily digest. Unsubscribe anytime.

We use your email only to send this newsletter.

Keep reading