The US Commerce Department on June 30 removed export controls on Anthropic's AI model "Claude Fable 5," clearing it for wider distribution after the startup resolved safety concerns raised by the Trump administration.
Jun 30, 2026 · Anthropic
US Lifts Export Controls on Claude Fable 5 — Global Redeploy Begins July 1
After the first-ever export controls applied directly to an AI model forced Anthropic to disable Fable 5 and Mythos 5 entirely, tighter safety classifiers cleared the way for a worldwide return.
Jun 12 → Jun 30
18 days under export controls before restrictions were lifted
99%+
block rate of the reported jailbreak after strengthened classifiers
1M
token context window (up to 128K output)
Senior-engineer benchmark score
Fable 5 vs prior frontier models — higher is better
Fable 5
General use · strengthened safeguards$10 / M input · $50 / M output
Mythos 5
For trusted partners · restricted version
From release to redeploy
JUN 9
Fable 5 & Mythos 5 released
→
JUN 12
Export controls invoked; models disabled
→
JUN 30
Controls lifted after safety fixes
→
JUL 1
Fable 5 redeploys globally
Praised
Called a "beast" — big gains in agentic coding and long-horizon problem solving; the 1M-token context a standout strength across partner benchmarks.
Criticized
"Slow and expensive." Doubts over how benchmarks reflect real use, the usage-credits pricing, and more false positives from tighter safeguards.
Continue reading The rest of this article is for AI News Blitz readers. Choose an option below to keep reading.
Already purchased? Sign in ✓ Signed in — this article isn’t included in your current plan.Unlocking the full article…