Artificial Intelligence
OpenAI extends Daybreak access to Ukraine
OpenAI said Ukraine's government would get access to Daybreak for cyber defence, while the article also cited Ukrainian incident figures and examples of earlier use of the tools elsewhere.
- OpenAI said it would give the Ukrainian government access to Daybreak to help protect civilian infrastructure from cyberattacks.
- The company said it was working with the Ministry of Digital Transformation, and that Ukrainian specialists would use the tools to find software weaknesses and handle suspicious activity.
- Sasha Baker said the programme was meant to help organisations improve defences and repair vulnerabilities, and said Ukraine needed support immediately.
- The article said CERT-UA handled 5,927 cyber incidents in 2025, which it put at 37.4% above the previous year.
- It also said Anthropic blocked accounts after Russia-linked users tried to use Claude for attacks on Ukraine and for military and intelligence purposes.
OpenAI said it would give the Ukrainian government access to Daybreak, which it described as a programme for legitimate cyber defence teams. It said the move was intended to help protect civilian infrastructure from cyberattacks.
The company said it was working with the Ministry of Digital Transformation. It said Ukrainian specialists would receive tools to scan software for weaknesses, investigate suspicious activity, confirm threats and generate fixes faster.
Sasha Baker, OpenAI's head of national security policy, said the tools would help public and private organisations strengthen their defences and repair vulnerabilities. She said Ukraine was already on the front line and needed support immediately.
The article said Ukraine was not the first place where OpenAI's cyber models had been used, citing France, Germany and Poland. It added that ENISA had used them to identify and patch vulnerabilities, and that CERT Polska had found six critical flaws in router software, prompting a vendor to issue updates.
Later, the article said CERT-UA handled 5,927 cyber incidents in 2025, which it said was 37.4% more than in 2024. It also said the SBU Cybersecurity Department had neutralised more than 16,000 Russian cyberattacks and critical incidents since the full-scale invasion began, and that Anthropic had blocked accounts after Russia-linked users tried to use Claude for attacks on Ukraine and for military and intelligence purposes.
Named in this story
People
- Sasha Baker
- OpenAI's head of national security policy who commented on the move
Companies
- OpenAI
- announced access to Daybreak for Ukraine
- Anthropic
- said it blocked accounts after misuse of its model
Organisations
- ENISA
- used OpenAI algorithms to find and patch software vulnerabilities
- CERT Polska
- used the models to find critical router flaws
- CERT-UA
- processed 5,927 cyber incidents in 2025
Governments and agencies
- Government of Ukraine
- received access to the programme
- Ministry of Digital Transformation
- worked with OpenAI on the rollout
- SBU Cybersecurity Department
- said it had neutralised more than 16,000 cyberattacks and critical incidents
Products and systems
- Daybreak
- the cybersecurity programme being made available
- Claude
- the model Russia-linked users tried to use
How the source tells it
The piece was mostly factual, but it used urgent defensive framing in the quote and added a mildly alarming statistics section about attacks in Ukraine.
- urgency a quoted remark framed the rollout as immediate support for Ukraine on the front line
- alarm the later incident statistics were presented as a sign of worsening pressure and concern