Anthropic in Talks With White House Over Fable 5's Return. Deal Remains Out of Reach

Anthropic in Talks With White House Over Fable 5's Return. Deal Remains Out of Reach

Ondřej Barták
Ondřej Barták
Entrepreneur and Programmer
17. 6. 2026
4 minutes reading · 11 views
Anthropic in Talks With White House Over Fable 5's Return. Deal Remains Out of Reach

Anthropic representatives are in Washington negotiating. Since last Friday, when the Trump administration forced the company to shut down the models Fable 5 and Mythos 5, both sides have been trying to find a way out. So far, without success.

Briefly, here is why this happened in the first place. On Friday, the government issued a directive barring foreign governments, companies, and numerous individuals—including foreign nationals working directly for Anthropic—from accessing Fable 5. The reason was concern that someone had found a way to "jailbreak" the model, meaning bypass its safety safeguards. As a result, the company withdrew Fable 5 and the more powerful Mythos 5 for all users. It is now trying to have the restrictions lifted.

A meeting that ended without an agreement

A key meeting took place at the Department of Commerce on Monday. Anthropic engineers met with government officials. The talks were led by the Commerce Department together with National Cyber Director Sean Cairncross.

The result? Nothing. According to people familiar with the talks, discussions continued without an agreement, putting even more pressure on both sides. It was not immediately clear which administration officials had actually attended the meeting. The entire discussion took place behind closed doors and was intended to remain confidential. Nor was it known exactly who from Anthropic had traveled to Washington.

Monday's meeting was not the first, however. The company's engineers have essentially been in constant contact with officials since Friday, when the directive took effect. According to a source familiar with the situation, Anthropic engineers traveled to Washington in person and have also spoken with the government virtually every day since the directive was issued.

Anthropic itself confirmed this in its first public statement since Friday. "This follows a series of virtual meetings we have held with the government since it first contacted us. Both sides are working toward a swift resolution," a company spokesperson said. "It is part of our long-standing commitment to work with the administration toward our shared goal of protecting America's critical infrastructure and maintaining America's lead in cyber defense."

What both sides are demanding

David Sacks, co-chair of the President's Council of Advisors on Science and Technology, described the administration's goal quite clearly. "The administration now hopes that Anthropic will fix the security issue, the export restriction will be lifted, and Fable will return to normal operations," he wrote on X. "The administration wants all of this to happen as soon as possible."

Sacks sharply criticized the company. He said the government and its "trusted partner" consider the jailbreak serious, yet Anthropic has still failed to provide a fix, even though company CEO Dario Amodei has repeatedly promised to put safety first. "It is frankly baffling that Anthropic does not want to comply with security requirements that it previously claimed were its highest priority," Sacks wrote.

Anthropic makes a different argument. It says the government presented only verbal evidence of a "potential narrow, non-universal jailbreak" that essentially consists of asking the model to read specific code and fix errors in it. The company investigated the jailbreak and says the same method can also bypass other models, such as OpenAI's GPT-5.5. Moreover, the vulnerabilities discovered appear "relatively simple."

According to Sacks, the current dispute is unrelated to previous tensions between Anthropic and the Pentagon. "Those who are trying to mistakenly link this action to previous disputes between the Department of Defense and Anthropic are wrong," he wrote.

Outside pressure is mounting

Outside voices have also begun weighing in on the negotiations. On Sunday, more than one hundred cybersecurity experts and company executives, including people from Adobe and Nvidia, signed an open letter to the administration. In it, they called for the export restriction to be lifted and for the government to commit to an "open, scientific, and transparent AI risk assessment process" in the future.

Their argument focuses on the security of the United States itself. According to the signatories, depriving America of its best defensive tools "without good reason" is dangerous, especially as adversaries rapidly close the gap. Chinese models are reportedly "only months" behind the best American ones. And time is working against everyone. The longer Fable 5 and Mythos 5 remain offline, the more pressure there is on both Anthropic and the White House to find a compromise. The company had previously discussed the capabilities of its models with the administration, so the two camps already know each other. But that did not move Monday's meeting any closer to an agreement.

The entire negotiation is also taking place at a sensitive time. Anthropic, led by Dario Amodei together with his sister Daniela as president, was recently valued at $965 billion and is preparing to go public. It has already filed paperwork for an initial public offering. The longer the dispute drags on, the more is at stake.

Sources: abcnews.com and theinformation.com

Advertisement

Content created with help from UpTier.

SEO and GEO on autopilot. UpTier’s multi-agent systems write and optimize content for search engines and AI answers.

Discover UpTier ↗

Category:AI
Did you enjoy this article?
Discover more interesting posts on our blog
Back to blog

Related posts

OpenAI introduces GPT-6.2 Astra and Sol for professional and everyday workOpenAI introduces GPT-6.2 Astra and Sol for professional and everyday work
OpenAI has introduced two models: Astra for complex professional tasks and Sol for everyday work. Both are available in ChatGPT Work, Codex and through the API.
3 min read
17. 11. 2026
Tencent releases Hy4 Flash with image support and a context window of up to 1 million tokensTencent releases Hy4 Flash with image support and a context window of up to 1 million tokens
The lighter Hy4 Flash model targets chat, documents and routine code edits. It accepts text and images, supports tool use and, according to Tencent, matched Claude Haiku 5.5 in tests of everyday tasks.
1 min read
16. 11. 2026
Companies Linked to Chinese Spy Invested in DeepSeek and Promised Banned Nvidia ChipsCompanies Linked to Chinese Spy Invested in DeepSeek and Promised Banned Nvidia Chips
Companies linked to an alleged Chinese spy joined DeepSeek’s latest funding round. In exchange for a stake, they reportedly promised banned Nvidia Blackwell chips.
3 min read
16. 11. 2026
Přihlaste se k odběru našeho newsletteru
Zůstaňte informováni o nejnovějších příspěvcích, exkluzivních nabídkách, a aktualizacích.
CodedTrip

Operated by CodedTrip LLC, USA.

YouTube
TikTok