How Anthropic may have talked itself into an AI export ban

Ars Technica AI · Clara Murray, Financial Times · 2026-06-22

A Financial Times quantitative analysis finds Anthropic uses AI risk and regulation language eight times more frequently than OpenAI in 2026, with critics arguing the company's persistent safety warnings directly contributed to the US government banning foreign nationals from accessing its newest models, Mythos and Fable.

Open original ↗

Appears in

US Government Export Control Directive Suspends Fable 5 and Mythos 5 for Foreign Nationals

Extraction

Topics: ai-regulationexport-controlsai-safety-messaginganthropicai-policy

Claims

Five in every 1,000 words Anthropic published in 2026 related to risk, regulation, or restrictions, compared to 0.6 per 1,000 for OpenAI.
The US government banned foreign nationals from accessing Anthropic's latest models, Mythos and Fable.
Some technologists attribute the export ban directly to Anthropic's own repeated public warnings about AI's societal risks.
The contrast in safety rhetoric between Anthropic and OpenAI has become politically charged in Washington.

Key quotes

Five in every 1,000 words used by Anthropic in 2026 related to risk, regulation, or restrictions... The equivalent figure for OpenAI and Sam Altman was eight times lower, at 0.6 words per 1,000.

Some technologists have blamed the decision on the $965 billion AI group's repeated warnings about AI's risk to society—particularly in relation to Mythos.