Key point: The White House removed Anthropic’s Fable model from the market with export controls after concerns about bypassed security safeguards, following failed intensive negotiations between government officials and CEO Amodei.
The Trump administration removed the AI model Fable from Anthropic after less than two days following concerns about bypassed security safeguards. An intense 24-hour dialogue between the White House and Anthropic CEO Dario Amodei ended with the imposition of export controls.
On Thursday, two days after the public release of the Fable model, Amazon CEO Andy Jassy conveyed concerns to the White House regarding the model’s security safeguards being bypassed. Amazon acted on a government request for feedback, noting that Amazon itself is an investor in Anthropic.
On Friday morning, the matter escalated to the highest level of the White House. Treasury Secretary Scott Bessent, White House Cyber Director Sean Cairncross, Chief of Staff Susie Wiles and other senior officials convened a meeting. Bessent participated remotely during a previously scheduled trip to Houston. Following the meeting, the administration attempted to reach Anthropic CEO Amodei but was told he was unavailable at a wellness retreat. Anthropic strongly disputed this claim, stating the statement was “absolutely false”. According to statements from Anthropic representatives, Amodei was contacted around midday and was available within approximately one hour and 15 minutes.
Amodei participated in three phone calls with a group of approximately six senior government officials, including Cairncross, Bessent and Commerce Secretary Howard Lutnick. Other participants included Under Secretary of Commerce for Industry and Security Jeffrey Kessler, White House Staff Secretary Will Scharf, Deputy Chief of Staff Richard Walters and Policy Assistant Walker Barrett. During the conversations, Amodei defended the model’s security safeguards, arguing that the observed bypass represented a specific method and did not pose the risk of a broader “jailbreak” that would make the model usable without any protections.
In a blog post following the imposition of export controls, Anthropic stated that “to date, no tester has found a universal jailbreak that could comprehensively bypass the model’s security safeguards”. The export controls resulted in the model’s withdrawal.
Source: www.politico.com · Published 14 June 2026
Lumi AI News — AI-assisted curation in accordance with Art. 50 EU AI Act. Paraphrase and classification via Lumi News Pipeline v1.7.1.