We all know that building a world-class AI model from scratch is an absolute nightmare—it takes an insane amount of money, computing power, and time. But some companies found a shortcut. The NSA, CISA, and FBI recently revealed that six Chinese AI firms—DeepSeek, Moonshot AI, Alibaba, MiniMax, StepFun, and Z.AI—have been unauthorizedly extracting capabilities from US frontier models.

Since late 2024, these companies used a technique called "distillation" to essentially siphon knowledge from models like Claude, GPT, Gemini, and Grok. By doing this on an industrial scale, they managed to slash their development time and costs significantly, likely with the Chinese government's blessing. It's a wild reminder of just how high the stakes are in the global AI arms race right now.