OpenAI Hands Ukraine Its Daybreak Vulnerability Hunter for Grid and Hospital Defense
OpenAI is giving Ukraine free access to Daybreak and GPT-5.6 Sol to defend the hospitals, power systems and telecom networks Russia attacks with both missiles and code.
OpenAI is giving Ukraine free access to Daybreak and GPT-5.6 Sol to defend the hospitals, power systems and telecom networks Russia attacks with both missiles and code.
OpenAI will give Ukraine access to its Daybreak cyber defense program, the company announced Sept. 23 on the sidelines of the UN General Assembly. It will work with the Ministry of Digital Transformation to protect civilian infrastructure.
The access will be free and include OpenAI's GPT-5.6 Sol model, the BBC reported. Daybreak reviews older software, investigates suspicious activity, validates vulnerabilities and tests fixes, OpenAI explained.
Frequently Asked Questions
What is OpenAI giving Ukraine?
Free access to Daybreak, its program of AI tools for authorized security work, plus the GPT-5.6 Sol model, delivered through Ukraine's Ministry of Digital Transformation, per OpenAI and the BBC.
What will Ukraine use it for?
Reviewing older software, investigating suspicious activity, validating vulnerabilities and testing fixes, per OpenAI. Consul General Dmytro Kushneruk listed incident response, threat triage, log analysis, system inventories and code analysis.
How big is the cyber threat to Ukraine?
CERT-UA, Ukraine's national incident response team, handled nearly 6,000 cyber incidents in 2025, including attacks on hospital systems, the energy sector and telecommunications.
Who else already uses OpenAI's cyber models?
Defenders in France, Germany and Poland, OpenAI said. ENISA and CERT Polska used them to find vulnerabilities that were later fixed.
Why would OpenAI do this?
OpenAI said Ukraine's defenders need support now. RUSI's Jamie MacColl told the BBC Western tech firms also gain valuable data and intelligence from an active conflict.
AI-generated summary, reviewed by an editor. More on our AI guidelines.
Read next — from the Lexicon
Below this point
7 more sections.
