analysis
Why the capability leap in Z.ai’s GLM-5.2 model reignites the open-source safety debate.
Open-weight AI safety risks are under fire as SaferAI finds Z.ai’s GLM-5.2 matches frontier capabilities but lacks vital safety guardrails.
Topic
Governance
Ultrathink's source-backed coverage of Governance. 1 story so far. Latest, on : Why the capability leap in Z.ai’s GLM-5.2 model reignites the open-source safety debate..