OpenAI and Poland’s NASK partner to advance AI safety and security

OpenAI signed an agreement with Poland's NASK National Research Institute to collaborate on evaluating capable AI systems, identifying vulnerabilities, and strengthening safety safeguards.
Key points
- NASK's AI Safety Research Center will participate in the research while serving as the Polish Artificial Intelligence Safety Institute
- Research areas include detecting synthetic AI-generated content, evaluating multilingual safety risks, and testing model safeguards
- The agreement follows an earlier collaboration where CERT Polska used OpenAI models to uncover six vulnerabilities in MikroTik RouterOS
Why it matters
Polish AI safety researchers gain formal access to OpenAI systems for safety testing, extending OpenAI's institutional safety agreements beyond the US and UK.
Sources
- OpenAI and Poland’s NASK partner to advance AI safety and security (OpenAI, Oct 23, 2026, 6:04 AM)