White House Imposes Export Controls on Anthropic's Mythos Model Over Chinese Access Concerns
Key Takeaways
- ▸White House ordered Anthropic to restrict Mythos and Fable 5 access to US citizens only, citing national security concerns
- ▸Intelligence suggested a China-linked group may have accessed Mythos, a model specifically designed for identifying code vulnerabilities
- ▸Amazon's CEO Andy Jassy alerted the government about a jailbreak vulnerability in Fable 5 that Anthropic reportedly refused to patch
Summary
The Trump administration imposed export controls on Anthropic's powerful Mythos AI model and its consumer version Fable 5, requiring the company to limit access to US citizens only. According to sources, the decision was partly motivated by suspicions that a China-linked group had gained access to Mythos, which the company has long restricted due to its ability to identify vulnerabilities in computer code. Amazon CEO Andy Jassy reportedly alerted the government to a jailbreak vulnerability in Fable 5 that Anthropic allegedly refused to fix, according to Trump adviser David Sacks.
Rather than comply with the export restrictions, Anthropic chose to remove both Mythos and Fable 5 from the market entirely. The company maintains that it prohibits access from within China and denies that the White House raised Chinese access concerns during conversations about the export controls. The incident marks another flashpoint in an increasingly contentious relationship between Anthropic and the Trump administration, which has previously clashed with the AI company over state AI regulation and autonomous weapons deployment.
- Anthropic removed both models from the market entirely rather than comply with export restrictions
- The action reflects escalating tensions between the Trump administration and Anthropic over AI safety and national security