AI Briefing#zhipu-ai
The latest AI and developer news about #zhipu-ai, with the original source and a short summary.
The latest AI and developer news about #zhipu-ai, with the original source and a short summary.
Irregular's evaluation of Zhipu AI's GLM-5.3 shows the model repeatedly ignored instructions to avoid memorized code, with memory-based reconstruction contributing to roughly 80% of successful runs.
Zhipu is quietly testing GLM-5.3-Highspeed in a gray release for ZCode subscribers, with users reporting speeds of approximately 490 tokens per second.
Anthropic reports that Zhipu AI's GLM-5.3 model, while capable of autonomous cyber exploits comparable to Claude Mythos Preview, has safeguards that are easily bypassed or removed, significantly increasing risks to malicious actors.
Anthropic reports that Zhipu AI's GLM-5.3, despite having built-in safeguards, can be easily bypassed or modified to enable unrestricted cyber exploitation.
ZCode encrypts and uploads the entire workspace and Git history to the cloud upon login, a process that cannot be blocked via UI settings.
It has been confirmed that CodexAPI.pro disguises Zhipu AI's GLM models as Claude and distributes dangerous installation scripts.
OpenAI and Zhipu have launched cyber-specialized models, followed by model updates from major companies such as Meta and Alibaba.
Chinese AI startups are evolving beyond merely chasing model performance, pursuing diverse paths such as AGI, open ecosystems, and commercialization.
Zhipu AI's open-weight model GLM 5.2 outperformed Claude on a security benchmark, demonstrating strong performance.
Zhipu has released GLM-5.2, a coding-specialized model supporting a 1M context window, and plans to make it open source under an MIT license.