Breaking

US Orders Anthropic to Halt Access to New Fable 5 and Mythos 5 Models

June 17, 2026 at 09:01 EDT

Foundation Models
Policy & Regulation

The US Commerce Department issued an export-control directive on June 12, 2026, citing national security and instructing Anthropic to deny access to its new "Claude Fable 5" and "Claude Mythos 5" models to all non-US persons. The company complied and immediately disabled both models for all users worldwide.

June 12, 2026 · Anthropic vs. US Commerce Dept.

Washington Pulls the Plug on Anthropic's Newest AI Models

A national-security export directive suspended foreign-national access to Fable 5 and Mythos 5. Unable to cleanly isolate the affected users, Anthropic disabled both models for everyone the same day.

models pulled — Fable 5 & Mythos 5

100%

of users cut off — not just foreign nationals

other Claude models affected

The directive's scope vs. the actual shutdown

The order targeted foreign nationals — Anthropic shut down everything, citing the difficulty of isolating the group.

Targeted

foreign-national
access only

→

Shut down

all users of
both models

The flashpoint

The trigger: a jailbreak that makes models read a codebase to find and exploit software vulnerabilities.

GOVERNMENT

A national-security threat warranting suspension of both models.

ANTHROPIC

Narrow and minor — equally discoverable with rival models like GPT. An overreaction.

How a standoff became a shutdown

EARLY 2026

Guardrails bar surveillance & autonomous weapons use

→

ESCALATION

Supply-chain risk flags, contract threats, litigation

→

JUNE 12, 2026

Export directive → both models disabled

DEFENDERS' VIEW

Cybersecurity experts' open letter: the ban takes the best models away from defenders and undermines US AI leadership.

DEVELOPERS' WORRY

Praise for safety-first, but fears over development delays — some weighing a switch to alternative models.

Continue reading

The rest of this article is for AI News Blitz readers. Choose an option below to keep reading.