- DeepSeek was recommended to be added to the US Entity List
- The company was accused of assisting Chinese military and intelligence
- White House avoided blacklisting companies ahead of Trump China visit
Despite claims from Anthropic that Chinese AI firm DeepSeek distilled its Claude model to improve their own models, and further evidence that DeepSeek supported Chinese military and intelligence operations, the US has held back on adding the firm to the Entity List.
Exclusive Reuters reporting, citing people familiar with the matter, claims the White House has avoided adding DeepSeek and more than 100 other Chinese firms to the blacklist to avoid inflaming tensions between the two countries any further.
The White House was recommended to add the firms to the Entity List by an interagency committee, but the administration avoided taking action ahead of President Donald Trump’s visit to China, where he met with Xi Jinping.
DeepSeek avoids US Entity List
Anthropic’s claims of distillation state that DeepSeek used over 16 million exchanges with 24,000 fraudulent accounts in order to distill the Claude model’s abilities.
“Distillation can be legitimate: AI labs use it to create smaller, cheaper models for their customers. But foreign labs that illicitly distill American models can remove safeguards, feeding model capabilities into their own military, intelligence, and surveillance systems,” Anthropic said in a statement on X. The claims made by Anthropic also target two other Chinese AI firms: Moonshot, and MiniMax.
Distillation can be legitimate: AI labs use it to create smaller, cheaper models for their customers.But foreign labs that illicitly distill American models can remove safeguards, feeding model capabilities into their own military, intelligence, and surveillance systems.February 23, 2026
Many US companies have turned to using DeepSeek as a cheaper alternative to US frontier models…
