OpenAI and Poland's NASK Partner on Groundbreaking AI Safety Research

OpenAI and Poland's NASK Partner on Groundbreaking AI Safety Research

In late October 2026, OpenAI officially announced a strategic partnership with Poland’s NASK (Naukowa i Akademicka Sieć Komputerowa) National Research Institute. The collaboration is focused on advancing AI safety, security, and technical research, marking one of the most concrete steps a major AI lab has taken to directly integrate government agencies into the process of securing frontier technology.

Through this agreement, NASK’s AI Safety Research Center—which serves as the Polish Artificial Intelligence Safety Institute (AISI)—will gain specialized access to OpenAI’s most advanced models and testing workflows.

A Proven Track Record with CERT Polska

The formal partnership builds upon a successful trial collaboration earlier in the year. In July 2026, CERT Polska, the national incident response team operating within NASK, was granted access to OpenAI’s cyber-capable models through a specialized channel called "Daybreak."

Using these advanced models, CERT Polska analysts successfully identified six vulnerabilities in MikroTik RouterOS. If left undiscovered, these flaws could have been chained together by malicious actors to take control of affected devices without authentication. The success of this early pilot proved the immense value of using cutting-edge AI for proactive vulnerability discovery and remediation, laying the groundwork for this broader alliance.

Why Poland? A Geopolitically Strategic Testing Ground

Poland occupies a highly sensitive geopolitical position, bordering active conflict zones and frequently serving as a primary target for digital attacks aimed at its energy grids, telecommunications, and government services. Because NASK has managed Poland's national internet domain and coordinated its cyber threat responses for years, it possesses deep, practical experience in monitoring threats at a national level.

By combining OpenAI’s technical capabilities with NASK’s frontline cybersecurity experience, the two organizations are creating a robust testing ground. Here, new defenses against AI-driven phishing, disinformation, and automated network attacks can be rapidly developed and validated.

Expanding Global Public-Private Cooperation

This partnership aligns with OpenAI's broader strategy of forging direct ties with national security and research institutions, moving beyond purely commercial enterprise relationships. The collaboration focuses on four core activities:

• Capability evaluation of increasingly advanced systems

• Risk assessment of AI deployment

• Safeguard testing to prevent misuse

• Coordinated incident reporting

For European policymakers, this alliance is a strong signal that American AI giants are willing to invest locally in safety cooperation, addressing the European Union's growing calls for responsible AI and digital sovereignty. Alongside its work with the UK AI Security Institute and the US CAISSI, OpenAI is working to ensure that the development of frontier models remains tightly aligned with international cybersecurity efforts.