Mythos Raises the Stakes: China Wants It Off the Table

Anthropic's Mythos AI model has become an unexpected flashpoint in US-China relations, with Beijing expressing serious concern that Washington could deploy the system as a geopolitical instrument ahead of a planned summit between President Donald Trump and Chinese President Xi Jinping, according to Bloomberg.

Chinese officials, according to people familiar with the matter, believe Mythos — Anthropic's most capable model to date — represents a qualitative shift in AI-enabled intelligence and economic analysis, one that could be directed against Chinese financial systems, supply chains, or military planning. The concern is less about what Mythos does today and more about what it signals: that the United States now holds an AI capability with no Chinese equivalent in reach or sophistication.

The anxiety lands at a delicate moment. Trump-Xi talks were already carrying the weight of unresolved tariff disputes, Taiwan tensions, and competing claims over semiconductor supply chains. Mythos adds a dimension that neither side's diplomats have formally addressed — and that neither side has the vocabulary to negotiate cleanly.

What makes this particular pressure point unusual is how little of it is visible. There are no missiles, no sanctions, no formal demarches. Just a model, running on servers somewhere in the United States, that Beijing has decided it cannot ignore.

The summit, if it happens, will now include a conversation nobody scheduled.