Fifty-four cybersecurity professionals, including Alex Stamos and Amazon CEO Andy Jassy, are demanding the White House reverse newly imposed export controls on two of Anthropic’s most advanced AI models, Fable 5 and Mythos 5.
Anthropic launched the heavily guardrailed Fable 5 on June 9. It was designed as a safer successor to Mythos, a model originating from the April 2026 Project Glasswing initiative known for identifying critical zero-day software vulnerabilities.
On June 12, the Commerce Department suspended foreign access. The order, triggered by a jailbreak incident exposing potential security flaws, arrived at 5:21 p.m. ET. Unable to implement nationality-based filters instantly, Anthropic disabled the models globally.
The signatories argue the decision creates a dangerous imbalance. They contend these tools are far more valuable for defending networks than for attacking them, and pulling them offline leaves defenders at a severe disadvantage.
Amazon’s involvement is significant. AWS is Anthropic’s primary cloud partner and a major investor. CEO Andy Jassy’s prior discussions with US officials reportedly contributed to scrutiny regarding the models’ security measures.
For the broader market, the timing is critical as Anthropic explores a potential IPO. The shutdown shifts the company’s narrative from major product launch to regulatory freeze in just 72 hours. White House adviser David Sacks has acknowledged a hope for a quick amendment to the restrictions.