White House science advisor Michael Kratsios accused China-based entities of conducting industrial-scale campaigns to steal frontier AI models from US companies using tens of thousands of surrogate accounts and distillation techniques. OpenAI warned that Chinese startup DeepSeek has been attempting to replicate ChatGPT's performance, while Anthropic identified DeepSeek and two other labs conducting campaigns to illicitly extract Claude's capabilities. The Trump administration plans countermeasures including information sharing with US AI companies, improved coordination, and accountability measures for foreign actors. The accusations come as the US and China remain locked in a race to build the most sophisticated AI technology, with AI becoming a major flashpoint in trade tensions between the two nations.
Story comments
Loading comments…