On June 12th, the US government issued an order to block foreign access to Anthropic’s latest AI models, Fable 5 and Mythos 5. The directive followed reports that Amazon security researchers identified a method to bypass safety filters, allowing the systems to generate information potentially useful for cyberattacks. Anthropic complied with the legal instruction by shutting down access to both models for all customers immediately after their launch on June 9th. The company stated it disagreed with the decision to recall a commercial product deployed to hundreds of millions of users based on a narrow potential jailbreak. This move adds to Anthropic’s existing tensions with US agencies, including a separate standoff with the Pentagon regarding autonomous weapons and surveillance technology. The shutdown effectively halts the global rollout of a system the company claimed exceeded the capabilities of any previously released model.
This incident highlights the growing friction between US regulatory bodies and private AI developers over safety and sovereignty. The White House’s intervention suggests a shift towards stricter oversight of large language models, potentially prioritising national security concerns over commercial deployment speed. For Anthropic, the recall damages credibility and sets a precedent where government directives can override corporate decisions on product availability. The situation also raises questions about the reliability of safety claims made by AI firms, as vulnerabilities were discovered shortly after public release. Furthermore, the ban may accelerate interest in non-American AI alternatives, as seen in the recent surge for Chinese models following similar restrictions on US technology.
* Anthropic has withdrawn access to Fable 5 and Mythos 5 globally following a White House order linked to security vulnerabilities.
* The shutdown marks a significant escalation in the conflict between US government regulators and major AI companies like Anthropic.
* The incident underscores the risk of rapid AI deployment and the potential for foreign nations to access restricted technology.
Stay ahead of AI. Get the most important stories delivered to your inbox — no spam, no noise.
