China Expresses Concern Over Anthropic Mythos AI Capabilities

The Government of China is expressing concern that Mythos, a frontier AI model developed by US-based Anthropic, could be utilized as an offensive weapon. This tension develops as Chinese President Xi Jinping prepares for a September 24 summit with US President Donald Trump. While Beijing aims for a "win-win" outcome to maintain a positive atmosphere before the meeting, it is drafting retaliatory measures, including sanctions and restricted-entity lists, in case the US targets Chinese AI firms.

In the US, officials remain concerned that Chinese companies use American AI models to train their own systems. Donald Trump is weighing additional AI controls while attempting to ensure the US does not lose its competitive edge to China. Treasury Secretary Scott Bessent has suggested unspecified sanctions if AI developers are found to have stolen intellectual property.

Anthropic CEO Dario Amodei has lobbied for tighter US restrictions on China's AI access and the export of high-end chips. These diplomatic frictions persist despite a 2025 truce concerning tariffs and rare earth materials.