Ukraine is significantly bolstering its digital shield by integrating Daybreak, the AI-powered cybersecurity tool from OpenAI. Deployed with the support of the Ukrainian Ministry of Digital Transformation, this technology leverages the GPT-5.6 Sol language model to automate the detection of critical vulnerabilities within aging IT infrastructures. Faced with an intensification of Russian cyberattacks—the volume of which surged by 37% in 2025—this tool not only identifies flaws but also validates the robustness of patches before they are deployed into production.
The Ukrainian landscape is marked by constant pressure from state-sponsored hacking groups, such as Sandworm, which was responsible for major sabotage including the attack on the operator Kyivstar. The adoption of this AI is part of a comprehensive digital defense strategy where rapid response is vital. Prior to Kyiv, security agencies in France, Germany, and Poland had already adopted these solutions, confirming a European trend toward relying on automation to counter persistent threats to essential networks.
This deployment marks a notable shift in OpenAI’s policy, following the lifting of initial restrictions early in 2024 that had prohibited the use of its technology for military purposes. While this collaboration is framed as humanitarian aid and support for critical infrastructure, some experts also point to the strategic interest for tech giants. By operating in an active conflict zone, these companies gain access to a real-world laboratory, allowing them to refine their models and gather valuable technological intelligence.
The implications of this automation go far beyond geopolitics, directly impacting the financial ecosystem. As cryptocurrency thefts reached new heights in 2025, with record losses tied to vulnerabilities in smart contracts, the ability of AI to audit computer code has become a major economic issue. The progress observed during competitions like the DARPA AI Cyber Challenge confirms that these tools could soon revolutionize on-chain protection, which currently remains heavily dependent on manual audits and bug bounty programs.