Question Everything. See who's reporting it.

China's AI firms disclose safety tests in just 3.6% of model releases

Source: Indian Express

The story in brief

A report by California-based SemiAnalysis found that only 3.6% of AI models released by nine Chinese companies between 2021 and September 15 had published safety-evaluation results. The report highlights global concerns over advanced AI risks, noting no major Chinese developer has disclosed comprehensive dangerous-capability tests for frontier models.

NEWXFree on Android and iPhone. Open in NEWX