Zhipu AI launches GLM-5.2 bug detector, signals China catching up
Technology
Chinese tech company Zhipu AI just dropped its GLM-5.2 model, which can spot software bugs and help with cybersecurity tasks, kind of like Anthropic's Mythos.
While it isn't quite as advanced as top US models from OpenAI or Anthropic, this launch shows China is catching up in the AI race.
US government flags open-weight GLM-5.2 risk
The US government isn't thrilled about GLM-5.2. They see powerful AI models that find security flaws as a national security risk, especially if they're widely available.
Unlike tightly controlled American models, GLM-5.2 is open-weight: anyone can download and run it on regular hardware, making it both accessible for innovation and potentially risky if misused by bad actors.