OpenAI teams up with Poland’s NASK to boost AI safety and security
ai-safety gpt-5 openai
| Source: OpenAI | Original article
OpenAI and Poland’s NASK have teamed up to boost AI safety, security and technical research, aiming to strengthen safeguards and build a more resilient digital ecosystem.
OpenAI has entered a formal partnership with Poland’s research institute NASK to deepen work on AI safety, security and technical research. The agreement grants NASK, and specifically CERT Polska, privileged access to OpenAI’s newly released GPT‑5.5 Cyber – a specialised version of the company’s language model built to detect, analyse and respond to cyber‑threats. Access is provided through OpenAI’s Trusted Access for Cyber programme, which is limited to vetted government partners.
The collaboration marks one of the first instances of a national cyber‑defence unit receiving a dedicated AI tool of this calibre. By integrating GPT‑5.5 Cyber into its workflow, CERT Polska expects faster vulnerability identification, more efficient malware analysis and accelerated incident response, bolstering Poland’s overall digital resilience. For OpenAI, the partnership extends its safety‑by‑design agenda beyond commercial products, offering a real‑world testbed for refining safeguards against misuse while showcasing the practical benefits of responsible AI deployment.
Looking ahead, the partnership will likely produce joint research outputs on model robustness, adversarial testing and threat‑intelligence sharing. Observers will watch how quickly the tool is operationalised within Poland’s national CERT, whether the model’s performance meets expectations in live environments, and if the collaboration spurs similar agreements across the EU. The outcome could influence broader regulatory discussions on AI‑driven cybersecurity and set a precedent for trusted‑partner programmes that balance innovation with stringent safety standards.
Sources
Back to AIPULSEN