OpenAI and poland's NASK forge alliance for AI safety and security research

OpenAI and Poland’s National Research Institute for Cyber Security (NASK) have entered into a partnership to enhance AI safety and security through collaborative research. The agreement aims to combine the expertise of both entities to better understand emerging risks and strengthen safeguards for advanced AI systems. The collaboration builds on OpenAI’s existing partnerships with global institutions such as the US Center for Advancing Standards and Innovation for Super Intelligence (CAISSI) and the UK AI Security Institute (AISI). NASK’s AI Safety Research Center, functioning as Poland’s AI Security Institute, will work alongside OpenAI to evaluate increasingly capable AI systems and identify potential vulnerabilities. In July 2026, CERT Polska, operating under NASK, accessed OpenAI’s cyber capable models through Daybreak, uncovering six vulnerabilities in MikroTik RouterOS. The vendor released fixes, which CERT Polska confirmed mitigated the identified threats.