OpenAI and Poland’s NASK partner on AI safety and security

A New Partnership for Safety

OpenAI and Poland’s NASK have officially announced a partnership focused on AI safety, security, and technical research. The collaboration is designed to strengthen safeguards and support a more resilient digital ecosystem. This move highlights a growing trend of AI developers working with state institutions to address the risks associated with large-scale models.

The announcement, which was published on October 23, 2026, frames the work as a way to improve digital resilience. While the partnership is significant, the initial statement did not provide specific operational details or a set timetable for the projects. It serves as a high-level commitment to collaborative research between a private tech leader and a public research body.

By focusing on both safety and security, the two entities aim to address the technical challenges of deploying AI in sensitive environments. The goal is to create a framework where AI tools can be used more reliably without compromising the integrity of digital infrastructure. This cooperation places the work at the intersection of model governance and national digital security.

Understanding NASK's Role

NASK is a Polish state research institute that operates under the supervision of the Ministry of Digital Affairs. It plays a critical role in the country's digital landscape by managing the national .pl domain registry. Additionally, it oversees the Nationwide Education Network, which is commonly known as OSE.

Beyond infrastructure, NASK is deeply involved in cybersecurity research and telecommunications services. The institute also works on digital education and the protection of citizens against harmful online content. Its broad mandate allows it to provide a unique perspective on how AI can impact public services and national safety.

The institute's expertise is well-established in the region. It handles a variety of tasks ranging from artificial intelligence research to the development of tools for online safety. This diverse portfolio makes it a logical partner for a company like OpenAI, which requires deep technical knowledge to secure its models against various threats.

Cybersecurity and Defense

A key component of NASK’s capabilities is its role in national cyberdefense. The institute houses CERT Polska, which was established in 1996 to perform CSIRT NASK tasks. These tasks include threat monitoring, providing early warnings, and responding to security incidents across the country.

CERT Polska is one of only three national-level computer security incident response teams in Poland. It is responsible for malware analysis, vulnerability assessment, and the development of defensive tools. This operational experience in defending against active cyber threats provides a practical foundation for the research being conducted with OpenAI.

The inclusion of these cybersecurity functions suggests that the partnership may look at how AI can be used to defend against cyberattacks. By combining OpenAI's model capabilities with NASK's incident response experience, the two may explore ways to automate threat detection. This type of cooperation is essential for protecting critical infrastructure from increasingly sophisticated digital threats.

Infrastructure and Education

NASK's influence extends into the educational and structural pillars of Polish society. The institute is actively involved in digital education and manages the Nationwide Education Network. This suggests that the partnership could eventually touch upon how AI is integrated into learning environments and public information systems.

Recent developments at NASK indicate a continued focus on infrastructure and online safety. For example, the institute recently released a report regarding who is responsible for content on various platforms. These activities show that NASK is thinking deeply about the governance of digital spaces and the responsibilities of platform operators.

The partnership with OpenAI arrives as NASK continues to manage its Infrastructure and Educational Projects Centre. By aligning with a major AI player, NASK can potentially bring more advanced technical research into its existing projects. This could help shape how the Polish government and its citizens interact with AI technologies in the future.

Looking Ahead at the Collaboration

While the current announcement is broad, the focus on technical research and security provides a clear direction. The partnership aims to create a more resilient digital ecosystem by addressing the risks inherent in AI development. As the work progresses, the specific projects and deliverables will likely become clearer to the public.

Because no pricing, market data, or commercial terms were supplied, the current focus remains on the research and safety aspects of the deal. The collaboration does not yet have a verified catalyst or a specific implementation date. However, the involvement of a state research institute suggests a long-term commitment to establishing safety standards.

The outcome of this partnership will likely influence how AI safety is approached in Europe. By combining private-sector innovation with state-led cybersecurity expertise, OpenAI and NASK are creating a model for international cooperation. This work will be important for anyone interested in how AI models are governed and secured on a national scale.