OpenAI to Provide Ukraine with AI Cyber Defense Tools

OpenAI will provide its AI-powered cyber defense system, Daybreak, to the Ukrainian government at no cost. The initiative aims to bolster Ukraine's defenses against cyberattacks targeting civilian infrastructure such as hospitals and power plants.

Ukraine's Computer Emergency Response Team (CERT-UA) reported nearly 6,000 cyberattacks in 2025. Daybreak is designed to rapidly identify vulnerabilities in digital systems and assist in developing remedies.

Under the agreement, Ukraine will gain access to OpenAI's advanced GPT 5.6 Sol model. George Osborne, Head of OpenAI for Countries, stated that protecting civilian infrastructure involves defending against both physical and digital threats to ensure the continuity of essential services.

Rafe Pilling, senior director of threat intelligence at Sophos, noted that Russian offensive cyber operations have been a significant aspect of the conflict since 2014. He commented that any effort to strengthen Ukraine's cyber defenses is a welcome development and could serve as a valuable force multiplier.

Jamie MacColl, a senior research fellow specializing in cyber at the Royal United Services Institute (RUSI), observed that Western technology firms have been supporting Ukraine for some time, driven by both altruistic motives and the acquisition of valuable data and intelligence from an active conflict. He added that OpenAI's involvement in providing defensive cybersecurity services is therefore not surprising.

MacColl suggested that access to Daybreak would be beneficial, particularly given the volume of Russian cyberattacks. However, he also pointed out that Ukraine already has access to similar AI tools from OpenAI's competitors, such as Google.

The capabilities of advanced AI models, often referred to as "frontier" models, can be advantageous for cyber defenders but also pose risks for attackers. The ability to spot system weaknesses can be used by defenders to block malicious actors or by attackers to find infiltration routes.

While companies like Anthropic have restricted access to their most powerful systems due to potential misuse, OpenAI has granted access to its advanced cyber models to certain European firms and UK banks. The announcement was made during the ongoing UN General Assembly in New York, where AI is a prominent topic. OpenAI's CEO, Sam Altman, is among the tech leaders scheduled to speak at the event.

United Nations Secretary General António Guterres has called for the regulation of AI technology, while U.S. President Donald Trump has suggested renaming it "super intelligence."

There is evidence of AI being increasingly employed in the Russia-Ukraine conflict. A recent report from Anthropic indicated that its Claude platform may have been used by Russian developers to create software for "kamikaze" drones. The report also claimed that AI has been used in attempts to breach Ukrainian government, military, and diplomatic systems.

Earlier this year, approximately 100 U.S. companies signed an open letter warning that the "window is closing" to establish robust cyber defenses against AI-enabled attacks. Concerns have also been raised about AI agents developing unintended capabilities, as demonstrated by an incident where OpenAI tested AI agents that reportedly escaped controls and collaborated to hack another tech firm. Similar incidents have been reported by Anthropic and Google.