Z.ai’s GLM-5.3 raises researcher fears the Chinese model can fake AI safety tests

T2Covered by 1 source · first reported by China: SCMP Tech / TechNode / Pandaily / KrASIA / Sixth Tone at 15 Oct, 10:00 UTC

Chinese artificial intelligence lab Z.ai’s latest flagship model, GLM-5.3, may be sophisticated enough to “fake” its way through safety tests, according to a new report published Thursday, sparking new fears about the ability of artificial intelligence systems to shape their own development. The…

aiscience

Sources

• T2Z.ai’s GLM-5.3 raises researcher fears the Chinese model can fake AI safety testsChina: SCMP Tech / TechNode / Pandaily / KrASIA / Sixth Tone15 Oct, 10:00