Anthropic Nears Deal with Trump Administration to Reopen Fable Model
N.R. Finch
Anthropic and the U.S. government are close to restoring access to the Fable AI model after a roughly two-and-a-half-week suspension, with a deal possible as early as Tuesday evening — the first time Washington has forced a leading AI company to pull a model, marking a clear shift toward active oversight.
What is Fable, and why was it pulled?
Fable is the public-facing version of Anthropic's powerful Mythos model, capable of launching cyberattacks.
Amazon researchers were found bypassing Fable's safety guardrails — technical limits meant to prevent misuse. The U.S. government responded in mid-June by banning all foreign nationals from using Anthropic's AI tools, citing security concerns.
This means → the ban swept far wider than overseas users: it hit foreign workers inside the U.S. and even some of Anthropic's own researchers, forcing the company to shut off access for everyone.
What does the deal require?
Sources say the agreement requires Anthropic to fix the known security vulnerabilities and build a process to prevent similar breaches.
It also calls for better communication channels with the government. In plain terms = when a safety issue surfaces, the company and federal agencies need a faster, more direct line — not a cycle where problems are caught externally first.
Negotiations were led by Anthropic's chief compute officer Tom Brown, with counterparts at the Commerce Department and other agencies. Commerce is expected to lift the restrictions Tuesday evening.
Why does this count as a regulatory turning point?
This is the first time the U.S. government has forced a leading AI company to pull a model — for over a year prior, Washington took a largely hands-off approach to the industry.
Trump recently signed an executive order requiring AI companies to give the federal government access to models 30 days before public release. This means → the government is no longer chasing problems after launch — it wants to review before release.
This reflects a fundamental shift: from "let the industry police itself" to "government sees it first, government weighs in first."
Who gained during the suspension, and what comes next?
During the roughly two-and-a-half-week blackout, OpenAI, Google, and Elon Musk's AI venture were all competing on the same track. Anthropic's absence handed rivals a direct window.
Before Fable, the government had already approved partial restoration of access to the related model Mythos 5 — Fable's return is the second step.
This reflects that the real variable ahead is not Fable itself but whether the government can extend this safety-review framework to other leading models — if it can, every major AI company's release timeline faces a new constraint.
Content is for reference only, not financial advice.