Why the US government shut down Anthropic’s latest Claude AI model

Here is a summary of the key points from the article on Anthropic’s Fable models:

Suspension of AI Models: On June 12, Anthropic suspended access to its latest AI models, Fable 5 and Mythos 5, following a government export control directive.
Mythos and Fable: Mythos is Anthropic’s most powerful model, and Fable is the same model with safeguards to prevent misuse, particularly in cybersecurity.
Conflict with Trump Administration: There has been escalating conflict between Anthropic and the Trump administration since early 2025, primarily over AI regulation and the Pentagon’s use of its models.
JAILBREAK Concerns: The directive was reportedly due to a jailbreak, a method to bypass Fable’s safeguards which the government feared could yield information useful for cyberattacks.
Amazon’s Involvement: The research behind the directive reportedly originated from engineers at Amazon, a key investor and competitor of Anthropic.
Opacity of AI: Large language models like Fable are complex and partly opaque even to their creators, which complicates regulation.
Future of AI Governance: Regulation is challenging due to the rapid development of AI, its unpredictability, and the lack of a global, participatory governance framework focused on mutual trust.

You may have missed