In June 2026, the U.S. government did something unprecedented: it made two of America's best AI labs restrict their most powerful models — while China gave frontier-class AI away for free. Here's why that might backfire.
Open-weight models from China (GLM, DeepSeek, Qwen, Kimi) have nearly closed the gap on everyday work at a fraction of the price. So when Washington gated Anthropic's and OpenAI's best models, it may have handed the rest of the world a reason to run on China's instead. The catch: you can switch off a closed model — you can't un-release an open one.
Some cinematic scenes are reconstructions of documented events. All quotes, dates, and figures are verified.
#AI #OpenSource #ChinaAI #OpenAI #Anthropic #techpolicy #artificialintelligence