Z.AI’s GLM-5.3 completes 3 of 13 cyberattack scenarios in Irregular’s simulated tests

Open-weight AI models like GLM-5.3 pose security risks due to their accessibility, potentially enabling misuse despite their capabilities.

The post Z.AI’s GLM-5.3 completes 3 of 13 cyberattack scenarios in Irregular’s simulated tests appeared first on Crypto Briefing.