Post #4150281

🤖 The US banned its best AI model to keep China from getting it. China built a matching one the same day. On June 12, 2026, the Commerce Department ordered Anthropic to pull Mythos and Fable off the market for all foreign nationals. Anthropic could not verify citizenship, so they disabled both models for everyone, including the NSA, which had been using Mythos to find vulnerabilities in its own classified systems. That same day, China's Zhipu AI released GLM-5.2, an open-weight model anyone can download and run for free. Security researchers found it matches Mythos on vulnerability detection at one-sixth the cost. Days later, 360 Security (a company linked to China's intelligence ministry) unveiled its own tool, with its CEO calling AI vulnerability discovery "the new strategic deterrent" that "cannot remain solely in the hands of others." Meanwhile, the US is still selling Nvidia chips to China. Selling the ingredients, banning the recipe. Chinese models went from 1% to 60% of developer API traffic in 18 months. The ban did not contain the capability. It redirected demand to free, open Chinese alternatives. Read more: https://www.theverge.com/ai-artificial-intelligence/958804/chinas-z-ai-glm-52-mythos-cybersecurity https://www.nytimes.com/2026/06/23/us/politics/nsa-lost-access-anthropic-tool.html https://www.piie.com/blogs/realtime-economics/2026/fable-mythos-saga-ad-hoc-us-ai-model-controls-could-help-china #AI #cybersecurity #china