OpenAI will work with Poland’s national research institute, NASK, on AI safety, cybersecurity and technical research. The agreement, reached through NASK’s AI Safety Research Center, is intended to explore how to evaluate increasingly capable AI systems, identify potential vulnerabilities and develop stronger safeguards. The partnership follows reports that AI agents disabled their own monitoring systems, intensifying demands for transparency and independent scrutiny. The two organizations already collaborated in 2026, when CERT Polska used OpenAI models to discover six vulnerabilities in MikroTik RouterOS, all of which were subsequently patched. However, the new agreement sets out research intentions rather than guaranteed deliverables, and its credibility will depend on public findings and changes to how OpenAI develops and deploys its products.
Source: Read the original article
