China Offers AI Talks With US, Warns of Retaliation

China said it is willing to hold talks with the United States on artificial intelligence, but vowed to retaliate if Washington takes actions against Chinese companies over accusations of systematic technology theft. The Commerce Ministry in Beijing responded Wednesday to U.S. security agencies' claims that China's top AI firms have extracted proprietary knowledge from American labs "at an industrial scale," characterizing the accusation as groundless. The statement marks a carefully calibrated diplomatic response in escalating tensions over AI model distillation—a technique used to extract knowledge from proprietary AI systems. The accused companies include DeepSeek, Moonshot AI, Alibaba, MiniMax, StepFun, and Z.AI. Beijing rejected the charges while simultaneously signaling openness to dialogue, emphasizing it stands ready for "constructive and professional talks" based on principles of "equality and mutual benefit." The timing of China's response follows a joint advisory issued by the NSA, CISA, and FBI on September 8, alleging that Chinese AI labs have conducted "aggressive, malicious, and targeted" distillation operations against U.S. frontier models since late 2024, extracting billions of tokens across millions of exchanges. U.S. officials and industry executives from OpenAI and Anthropic have repeatedly leveled similar accusations. China's dual approach—offering talks while threatening countermeasures—reflects the complexity of U.S.-China AI competition at a moment when performance gaps between leading models from both countries have narrowed significantly. The standoff comes ahead of a scheduled summit between President Trump and Chinese President Xi Jinping on September 24, with preliminary indications that AI governance and safety may feature prominently in discussions.