30 September 2026

OpenAI gives cyber defence tools to Ukraine

Bbc | Zoe Kleinman

OpenAI will share its Daybreak AI cyber defence system for free with the Ukraine government to protect civilian infrastructure, such as hospitals and power plants, from cyber-attacks. Under the agreement, Ukraine will receive access to OpenAI's GPT 5.6 Sol model as it counters nearly 6,000 cyber-attacks recorded in 2025 by CERT-UA.

Russian offensive cyber operations have been a key component of Moscow's aggression since 2014, prompting tech firms to supply defensive tools while gathering valuable conflict data. Analysts note that frontier AI models possess dual-use capabilities, serving defenders while simultaneously posing risks when exploited by malicious actors for reconnaissance and infrastructure infiltration. Meanwhile, tech executives gathered at the UN General Assembly amid growing warnings from industry leaders regarding the rapid proliferation of autonomous cyber threats.

Comment

The integration of frontier artificial intelligence models like GPT 5.6 Sol into active theater defence infrastructure shifts the burden of response from manual patch management to automated threat mitigation. While tools such as Daybreak accelerate weakness identification for systems protecting civilian infrastructure, they operate within a contested digital environment where adversarial actors simultaneously leverage equivalent capabilities for automated reconnaissance. This technological parity transforms national cyber resilience into a continuous algorithmic contest rather than a static defensive fortification.

Strategic Question for Discussion
How does the deployment of proprietary defense models like Daybreak alter the operational autonomy of national incident response teams when managing large-scale infrastructure threats?
The reliance on external commercial systems introduces complex dependencies between sovereign defense operations and private technology providers. This dynamic suggests that national response efficacy will increasingly hinge on the speed of proprietary updates rather than indigenous capability development.
Share your assessment in the comments below.