OpenAI announced on Thursday that it will supply Ukraine with its latest GPT‑5.6 Sol model, a sophisticated generative‑AI system designed to aid cyber defence operations.

AI tools aimed at repelling cyber attacks

The agreement, reported by the BBC, gives Ukrainian cyber units access to a suite of AI‑driven capabilities, including automated threat analysis, real‑time code‑generation for patching vulnerabilities and predictive modelling of Russian hacking patterns.

OpenAI officials said the technology is a rival to Anthropic’s Mythos and Fable platforms, which have already been trialled by several NATO allies.

OpenAI GPT‑5.6 Sol interface displayed on a laptop screen

Why the move matters for the war effort

Since the invasion began, Kyiv has relied heavily on Western cyber assistance to protect critical infrastructure, from power grids to government communications. Experts estimate that Russian actors have launched thousands of attacks each month, many aimed at disabling civilian services.

By integrating a large‑language model that can parse massive codebases and generate defensive scripts within seconds, Ukrainian teams hope to shrink the response window that previously stretched from hours to minutes.

Broader implications for AI in conflict

The deployment underscores a growing trend of AI being weaponised beyond the battlefield. Nations are increasingly embedding generative models in intelligence, surveillance and logistics, blurring the line between conventional weapons and software tools.

Policy analysts warn that the rapid diffusion of such technology could spark a new arms race, as adversaries seek comparable AI capabilities to launch offensive cyber operations.

International reaction and next steps

Western officials welcomed the partnership, describing it as a "significant boost" to Ukraine’s digital defences. The United States Department of Defence, while not commenting on the specifics, reiterated its commitment to supporting allied cyber resilience.

OpenAI, which has faced criticism over the potential misuse of its models, said it will impose strict usage controls and monitoring to prevent the technology from being repurposed for offensive actions.

Implementation is slated to begin within weeks, with training sessions for Ukrainian cyber engineers planned in partnership with NATO’s Cooperative Cyber Defence Centre of Excellence.

The move arrives as the European Union debates new regulations on the export of high‑risk AI systems, a debate that may shape how future AI‑enabled tools are shared in conflict zones.

For now, Kyiv’s cyber command aims to integrate GPT‑5.6 Sol into its existing workflows before the winter, when heightened cyber activity is expected.