A new report from Semafor indicates that the White House's decision to impose export restrictions on Anthropic's Mythos was partly motivated by concerns that a group linked to China may have accessed the advanced AI model.
Potential National Security Risk
If the Chinese government indeed gained access to Mythos 5 or Fable 5, it could pose a serious national security threat. The government might attempt to reverse engineer the model through distillation, a process where a student AI is trained on a more advanced model to mimic its behavior.
White House and Adviser Responses
The White House has not confirmed the report. Trump adviser David Sacks posted on X about the models' alleged jailbreak vulnerability, which Anthropic denies, without mentioning China. Anthropic has not commented, but a spokesperson told Semafor that China was not raised during export control discussions.
If Mythos was accessed by China, it would not be the first breach of Anthropic's most powerful model. The company claims Mythos is too dangerous for public use, yet a Discord group reportedly had access for two weeks before Anthropic discovered and terminated it.
