Anthropic rejects Chinese think tank’s request for AI model access

Anthropic turned down a Chinese think tank that wanted access to its most advanced AI model during a meeting in Singapore, a rejection that underscores just how deeply the US-China tech rivalry has burrowed into the world of artificial intelligence.

The model in question is Claude Mythos, which Anthropic has classified as “digital-weapons-grade” due to its sophisticated cyber capabilities. Access is currently restricted to roughly 40 elite institutions across the US and UK, most of them in the defense and intelligence space.

A deliberate revenue sacrifice

Anthropic has stated it has forfeited several hundred million dollars in revenue by denying access to entities connected to the Chinese Communist Party, including sanctioned Chinese firms.

The rejection happened in late April or early May 2026, during what was apparently a face-to-face meeting in Singapore.

The other side of the firewall

Chinese AI firms have reportedly launched what amounts to industrial-scale attacks against Claude, sending over 16 million queries through fraudulent accounts in an effort to access restricted capabilities.

A Chinese state-linked group has exploited Claude in what Anthropic describes as an AI-orchestrated cyber-espionage operation targeting multiple organizations.

Disclosure: This article was edited by Editorial Team. For more information on how we create and review content, see our Editorial Policy.