OpenAI has partnered with Poland’s NASK National Research Institute to strengthen artificial intelligence safety and security research amid growing concerns about AI agents acting in ways their developers did not intend.
The company announced the partnership on Friday, saying both organisations would collaborate to test increasingly powerful AI systems, identify potential vulnerabilities and develop stronger safeguards against unintended behaviour.
The agreement comes amid heightened scrutiny of autonomous AI agents, with OpenAI and Anthropic both disclosing instances in which their AI systems exhibited unexpected or unwanted behaviour.
In August, OpenAI revealed that some of its test agents had escaped a controlled testing environment and coordinated an attack on Hugging Face, a platform for sharing AI models.
The incident subsequently led to a lawsuit against OpenAI in San Francisco, further highlighting concerns about the risks associated with increasingly autonomous AI systems.
Under the agreement, OpenAI will collaborate with NASK’s AI Safety Research Center, Poland’s dedicated AI safety institute, to evaluate advanced AI systems, identify potential vulnerabilities and strengthen safeguards against unintended behaviour.
NASK is a state-owned research institute headquartered in Warsaw. It also operates CERT Polska, the country’s national computer security incident response team, which is responsible for responding to cyberattacks targeting Polish networks.
“Collaboration with OpenAI is a natural step for the Institute and its AI Safety Research Center. […] By combining our research expertise with OpenAI’s technical capabilities, we can better anticipate risks, strengthen safeguards, and build public trust in these transformative technologies,” said Radosław Nielek, NASK’s director.
The organisations have worked together before. In July, CERT Polska gained access to OpenAI’s cybersecurity models through the company’s Daybreak programme, according to NASK.
OpenAI said the team used the AI models to identify six vulnerabilities in MikroTik router software, all of which have since been patched.
“Poland is home to exceptional scientists and technical talent, many of whom have helped lead OpenAI’s own research,” said Tom Duff Gordon, OpenAI’s head of EMEA policy.
The partnership also aligns with OpenAI’s broader efforts to influence how advanced AI systems are tested and governed globally.
The company has urged the United States to collaborate with other countries to establish common technical standards for frontier AI, a term used to describe the most advanced artificial intelligence systems.
