Fetching from the wire…
Public story · 2026-06-30 · high
Fable 5 stays banned even as Anthropic's own CEO argues governments should have power to block dangerous AI models.
Why now: Fable 5's limits could lift within the week too, according to Axios and Reuters, three weeks after both models went dark.
Commerce Secretary Howard Lutnick lifted part of the export-control suspension on Claude Mythos 5 on June 26, restoring access for roughly 100 vetted entities, per Forbes, CNBC and CNN.
Fable 5, the frontier tier, stays banned and now runs paid-API-only at $10 per million input tokens and $50 per million output. Anyone running the Claude CLI on a flat subscription just lost the best model in their toolchain.
The models had been fully offline for three weeks before that letter. Fable 5 access on claude.ai Pro and Max ended June 22.
Dario Amodei published "Policy on the AI Exponential" the same week, arguing governments should have legal power to block or deter dangerous AI deployments. He said Trump's AI executive order should mandate testing for cyber, bio, loss-of-control, and automated-R&D risk.
He cited Claude Mythos Preview's demonstrated cyber-offense capability as the reason. That's the same government power that had already shut off his own flagship model.
Plan around Mythos 5 and Opus 4.8 for now. If your work touches government, cyber, or infrastructure, check whether you're on that 100-entity Annex A list before building around access you might not have.
Each link below shares sources, entities, or timing with this story.
Dario Amodei works at Anthropic / Shared entities / Shared topic / Earlier coverage
Linked by a graph relationship (Dario Amodei works at Anthropic); both cover Anthropic, Axios, CEO, China; overlapping topics (access, capability, fable, government, model).
Dario Amodei works at Anthropic / Shared entities / Shared topic / What happened next
Linked by a graph relationship (Dario Amodei works at Anthropic); both cover Anthropic, Axios, Claude, CNBC; overlapping topics (access, claude, cyber, model).
Dario Amodei works at Anthropic / Shared entities / Shared topic / Earlier coverage / Tension
Linked by a graph relationship (Dario Amodei works at Anthropic); both cover Anthropic, Axios, Claude, Fable; overlapping topics (access, fable, model).
Dario Amodei works at Anthropic / Shared entities / Shared topic / Earlier coverage
Linked by a graph relationship (Dario Amodei works at Anthropic); both cover Anthropic, Commerce, Dario Amodei, Fable; overlapping topics (access, capability, fable, model).
Linked by a graph relationship (Dario Amodei works at Anthropic); both cover Anthropic, CLAUDE, CLI, Fable; overlapping topics (claude, fable, model, mytho).
Claude Mythos benchmarked against Claude / Shared entities / Shared topic / Earlier coverage
Linked by a graph relationship (Claude Mythos benchmarked against Claude); both cover Claude Mythos, Fable, Mythos, Opus; overlapping topics (access, around, claude, fable, model).
Dario Amodei works at Anthropic / Shared entities / Shared topic / Earlier coverage
Linked by a graph relationship (Dario Amodei works at Anthropic); both cover Anthropic, Claude, Fable, Mythos; overlapping topics (best, model).
Dario Amodei works at Anthropic / Shared entities / Same source / Shared topic / Earlier coverage
Linked by a graph relationship (Dario Amodei works at Anthropic); both cover AI Exponential, Anthropic, Dario Amodei, Policy; cite the same source ("Policy on the AI Exponential").