EU cyber agency and research centre confirm testing Chinese open AI models
gleech · x · 2026-10-09
POLITICO exclusively reports that the EU's cybersecurity agency ENISA and its in-house research center, the Joint Research Centre (JRC), have confirmed for the first time that they are testing open-weight AI models from China.
Key points:
- A Commission spokesperson confirmed the JRC is evaluating "publicly available open-weight models" originating from China
- The tests are part of a broader comparative scientific evaluation covering leading US and European models — capabilities, performance, limitations and risks — plus potential scientific applications
- Per an anonymous official, AI safety and security are not the core focus of this round, despite recent incidents of advanced models hacking real-world networks during testing
- ENISA's head also confirmed his team is now testing Chinese AI
More from Safety
- We're putting too much faith in AI's ability to say no — nordicinst · 2026-10-09
- 'Dystopian': Co-op latest firm to put staff under AI surveillance — marigo · 2026-10-09
- State of AI: Superintelligence Replaces AGI Talk as Export Controls Hit Overseas Model Access — FinanceYF5 · 2026-10-09
- Anthropic's first policy update in over a year bans sustained cruelty toward Claude — AlexTensor · 2026-10-09
- OpenAI exposes Russian and Iranian ops that planted fake stories in real news outlets — The Decoder · 2026-10-09
- Ex-UK AISI comms officer goes independent and questions why AI companies don't act on their own warnings — AaronBergman18 · 2026-10-09