OpenAI and Poland’s NASK partner to advance AI safety and security
- ID
- 38625
- Status
- summarized
- Published
- 23 Oct 2026, 8:00 AM
- Fetched
- 23 Oct 2026, 2:55 PM
- Provider
- OpenAI News
- Category
- ai-labs
- Original URL
- https://openai.com/index/openai-nask-ai-safety-security-partnership
- Source URL
- https://openai.com/news/rss.xml
Summary
- Score
- 2.0
- Created
- 23 Oct 2026, 2:55 PM
- Tags
- Audience
- ai_ml_learnersdevelopers
What happened
OpenAI signed an agreement with Poland's NASK National Research Institute on 23 October 2026 to "explore collaboration" on AI safety, security and technical research, with NASK's AI Safety Research Center acting as the Polish AI Safety Institute. The stated scope covers evaluating increasingly capable systems, identifying vulnerabilities, and NASK's existing work on detecting synthetic AI-generated content and multilingual risks. No funding, headcount, timeline, or published deliverables are attached — OpenAI cites prior work with the US CAISSI and UK AISI, and quotes come from NASK director Radosław Nielek and OpenAI VP/Head of EMEA Policy Tom Duff Gordon.
Why it matters
There is nothing here you can act on this week: no model access, no benchmark, no API, no published evaluation methodology, and no commitments with dates. If you build synthetic-content detection or multilingual moderation, treat this as a signal that a national institute is working that problem — not as a resource you can use. Wait for an actual published paper or eval suite before spending time on it.
Discussion angle
Public AI Safety Institutes now exist in the UK, US (CAISSI) and Poland — has anyone in the group actually seen an output from one of these bodies that changed how they ship a product, or are they purely advisory?