A Chinese chatbot has again sparked concern in Washington. Moonshot AI, a Beijing-based firm, released its latest model, Kimi K3, on Friday, claiming it had “consistently outperform[ed]” every other model except the latest from U.S. rivals OpenAI and Anthropic. Chinese tech giant Alibaba followed two days later with similar claims about its soon-to-be-released model, Qwen 3.8. The new models have reignited questions about whether the United States still leads China in the AI race. Kimi became so popular that Moonshot suspended new subscriptions within 48 hours of launching, citing an overload of hardware used to power its models. Analysts say this betrays compute constraints Chinese AI developers face, as U.S. companies dominate advanced semiconductor chip production and the U.S. government has cut off China’s access through export controls. Ryan Fedasiuk, a fellow at the American Enterprise Institute, called Kimi’s launch a “second DeepSeek moment” but noted it “also illustrated some of the serious structural constraints facing China’s AI industry.” OpenAI and Anthropic have accused Chinese AI companies of using advanced U.S. models to train their own through a process called distillation, which U.S. officials allege is used for “illicit purposes” to steal proprietary technology. Matt Sheehan of the Carnegie Endowment said distillation helps make up for China’s compute deficit, adding, “If they should be 40