US AI crackdown threat sparks backlash from major tech firms

Twenty-five Silicon Valley firms, including Nvidia, Microsoft, and Meta, warned Washington against a US AI crackdown on Chinese open-weight AI models. In an open letter dated July 24, they argued that restricting these systems would backfire on US competitiveness. The pushback is tied to rapid progress in open-weight models, highlighted by Moonshot AI’s Kimi K3 (released around July 17). Kimi K3 reportedly matches or beats major US models on coding benchmarks and is priced about 40% lower. Other Chinese open-weight players such as DeepSeek and Alibaba’s Qwen are also gaining traction. Supporters cite market scale: Martin Casado (Andreessen Horowitz) said roughly 80% of startup pitches using open-source AI stacks now use Chinese open models. Smaller firms echoed the concern via the Little Tech Association (nearly 200 companies), arguing that an access ban would not stop global usage—developers outside the US (Europe, India, Southeast Asia, and elsewhere) would continue using the models. Investors should also note the political signaling. On July 17, Xi Jinping promoted China’s leadership in open-source AI and proposed international cooperation around the technology. With both large incumbents and startups aligned against the US AI crackdown, policymakers may face harder pressure to move quickly.