OpenAI and Poland’s NASK National Research Institute have signed an agreement to cooperate on artificial intelligence (AI) safety, security and technical research. The partnership aims to combine expertise from both organizations to better understand emerging risks, strengthen safeguards, and explore how advanced AI can support a more secure and resilient digital ecosystem.
Why the cooperation matters
As AI systems become more capable, international collaboration between developers and public research institutions is increasingly necessary to understand risks and develop stronger protections. OpenAI’s prior partnerships — including work with the US Center for Advancing Standards and Innovation for Super Intelligence (CAISSI) and the UK AI Security Institute (AISI) — have already helped harden safeguards, identify and remediate security vulnerabilities in AI agents, and advance frontier AI safety research.
Under the agreement, NASK’s AI Safety Research Center — acting as the Polish Artificial Intelligence Safety Institute (AISI) — and OpenAI will work together to evaluate increasingly capable AI systems, identify potential vulnerabilities and strengthen protective measures.
Linking international standards with national experience
OpenAI has urged the United States and other countries to cooperate on international technical standards for frontier AI, leveraging the expertise of AI safety institutes and standards bodies. Shared approaches to evaluating capabilities, assessing risks, testing safeguards and reporting incidents can help countries compare evidence and respond to cross-border risks. The collaboration with NASK illustrates the value of bringing national-level expertise into that broader effort.
NASK’s role and earlier cooperation with OpenAI
NASK’s expertise rests on extensive AI safety research applied to protecting public services and Poland’s digital information ecosystem. The agreement stresses that without national partners like NASK, global AI safety standards risk becoming disconnected from the contexts in which AI is actually deployed.
The agreement also builds on OpenAI’s existing work with Polish cyber defenders. In July, CERT Polska — which operates within NASK — gained access to OpenAI’s cyber-capable models via Daybreak. Using OpenAI models, the CERT Polska team helped discover six vulnerabilities in MikroTik RouterOS, including flaws that could be chained to take control of affected devices without authentication. The vendor issued fixes, which CERT Polska confirmed prevented the attacks it had observed.
Implications
The NASK–OpenAI partnership is expected to contribute to a deeper understanding of frontier AI safety and to the development of practical defenses, particularly for national-level applications and critical infrastructure. It also demonstrates how collaboration between technology companies and national research institutes can uncover and address real-world vulnerabilities.
