Trump Administration Allows Anthropic to Release Mythos to Select US Organizations
Get the Tech newsletter
Daily tech — startups, AI labs, chips, the launches that shape the next decade. Free.
- Commerce Secretary Howard Lutnick sent a letter to Anthropic cofounder Tom Brown permitting access to Claude Mythos 5 for 100+ US organizations—including large corporations and government agencies—after determining "appropriate safeguards are in place."
- Claude Fable 5, the consumer-facing version of Mythos, remains restricted with no broader rollout approved; Anthropic says weekend discussions with the White House over restoring it are ongoing.
- The White House's June 12 directive was triggered by concerns Anthropic had granted Mythos access to a South Korean telecom firm with alleged China ties, compounded by separate warnings from Amazon and the NSA that Fable 5 could be jailbroken.
- Organizations approved for Mythos 5 — and Anthropic itself — may now let their foreign national employees access the model, quietly reversing the original directive's restrictions that had covered even non-citizens living and working in the US.
- OpenAI announced Friday it is delaying its GPT 5.6 models at the Trump administration's request, a parallel development showing the Mythos fight is reshaping how frontier labs coordinate releases with Washington.
- Dean Ball, OpenAI's head of strategic futures and a former White House AI adviser, wrote June 16 that the spat proves frontier AI developers "need an explicit green light from the government now."
Why it matters: The partial reinstatement resolves a roughly two-week standoff but leaves Fable 5 still gated. With OpenAI now also delaying GPT 5.6 at the government's request, Commerce is asserting a de facto veto over frontier model releases — and Anthropic's earlier lawsuit over a military-use supply chain designation shows this is one battle in a broader AI policy war, not a one-off.
Ask SkimNews
