What's happening?
The European Union's cybersecurity agency ENISA and the European Commission's science service, the Joint Research Centre (JRC), have started testing open-weight AI models originating from China. Politico reported the work, and three EU officials confirmed it.
Why it matters
Open-weight models are published with their weights available, meaning anyone can download, run and study them. That makes evaluating them for security, data protection and dependency risks increasingly important.
First public acknowledgement
According to the report, this is the first time either body has spoken about the testing effort. The question of how Chinese-origin AI models should be handled from a security and regulatory standpoint is becoming more prominent within the EU.
What remains unclear
The scope of the tests, which specific models are being examined and how the results will be used are not detailed in the source material.
