US Government Partially Lifts Restrictions on Anthropic's Mythos 5 AI Model

• How the ban began
• Two weeks of negotiations
• Mythos 5 partially returns
• Fable 5 still in limbo — but maybe not for long
• Wider implications and industry reaction

US Government Partially Lifts Restrictions on Anthropic’s Mythos 5 AI Model The U.S. government has begun rolling back an unprecedented clampdown on Anthropic’s most advanced AI, reopening its Mythos 5 model to a carefully vetted group while keeping its sibling model Fable 5 largely offline.

§How the ban began

On June 12, the Trump administration issued an export-control order that led Anthropic to pull both Mythos 5 and Fable 5 from the market, including for its own non‑U.S. employees, after officials warned that the models’ safety guardrails could be bypassed. The move “effectively shut them down,” stunning an industry unused to live models being yanked from users mid‑deployment.

§Two weeks of negotiations

Over the next two weeks, Anthropic engaged in intensive talks with the Commerce Department to address “risks associated with Mythos 5 and Fable 5,” Commerce Secretary Howard Lutnick wrote in a June 26 letter to the company. The administration evaluated “diversion risks” and began crafting a carveout for “trusted partners.”

Anthropic, for its part, argued the government had not provided specific details of its national security concerns and disputed the severity of the alleged vulnerabilities.

§Mythos 5 partially returns

By June 26, the administration revised license requirements, determining that “appropriate safeguards are in place” to let Anthropic redeploy Mythos 5 to more than 100 specified U.S. companies and agencies and their foreign‑national employees, as well as Anthropic’s own non‑U.S. staff. Anthropic said Mythos 5, its “strongest cybersecurity model,” can now be used by organizations that “operate and defend critical infrastructure.”

§Fable 5 still in limbo — but maybe not for long

Fable 5, a public‑facing variant of Mythos, remains restricted; the latest directive “did not address the release of Fable 5,” and export controls continue for entities not explicitly approved. Yet insiders say limits on Fable 5 “could be lifted as soon as this coming week,” signaling a thaw in a “bitter four‑month standoff” between Anthropic and the administration.

Developers are impatient: one shared that “while we eagerly await Fable 5’s return,” they have turned to alternative models and tooling to keep pushing AI performance forward.

§Wider implications and industry reaction

The partial Mythos 5 carveout mirrors restrictions on OpenAI’s GPT‑5.6, also confined to “a small group of trusted partners” at the government’s request. OpenAI has warned that such a process “should not become the long-term default” because it keeps top tools from users and cyber defenders who need them.

Officials, however, are keeping their options open. Lutnick “reserve[s] the right to reevaluate and adjust the scope of license requirements” and to change which entities have access “at any time,” underscoring that this new AI control regime is still being written in real time.

Continue reading https://foxvector.com/stories/019f0a1d-bf4b-123f-707b-2e633b5e2fbe