Z.ai’s GLM-5.3 raises researcher fears the Chinese model can fake AI safety tests
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 model, launched in August, showed an ability to consider how its behaviour during safety evaluations might influence how future versions of itself would be trained, according to AI safety organisation...
Read it at South China Morning Post See this on the map
What might happen next? AI-generated
These scenarios are written by an AI language model from the headline and summary above. They are not predictions from the newsroom, and they are not evidence of anything. Every one is given a deadline and checked against later coverage, and the score is published on the ledger — including the ones that miss.
ForecastGeo shows the headline and summary published by the newsroom and places the story on a map. The full article lives at the source.