Core Summary

The U.S. government issued a 90-minute emergency directive to AI company Anthropic on June 13, demanding the immediate removal of its flagship Fable 5 model. The sudden storm was triggered by a security research report from Amazon that revealed critical vulnerabilities in Fable 5. Anthropic CEO Dario Amodei was forced to comply after participating in three urgent calls with administration officials, marking the most controversial government intervention in U.S. AI regulatory history.

Event Details

According to multiple media reports, the crisis was ignited by findings from Amazon’s security team. Amazon submitted a report to the White House identifying exploitable jailbreak vulnerabilities in the Fable 5 model. AI Czar David Sacks later confirmed that Amodei refused to either fix the vulnerabilities or deactivate the model when instructed, which directly triggered the export control measures.

The timeline shows that Anthropic received a call from the government on Friday with only 90 minutes to respond and complete the model takedown. Amodei personally participated in three emergency calls from senior administration officials. Anthropic was ultimately forced to disable both Fable 5 and Mythos 5 models for all customers to ensure compliance.

Notably, more than five companies reportedly pressured the government to push for the regulatory action. Amazon’s CEO had warned the government about safety risks in Anthropic’s models before the crackdown, revealing how competition among tech giants has profoundly influenced AI regulatory policy.

Panoramic Analysis

This event reveals a disturbing trend in the AI industry: commercial competition is shaping national security policy in unprecedented ways. When a single company’s security research can trigger emergency government action against a competitor, and a 90-minute response window represents the extreme compression of corporate autonomy, this pattern could fundamentally alter the innovation ecosystem if it becomes precedent.

From a regulatory perspective, the U.S. government is treating the most advanced AI models as strategic assets equivalent to weapons. Export controls now target not just foreign entities but directly intervene in domestic companies’ product deployment. This “domestic-foreign policy” approach blurs the traditional boundaries between commercial freedom and national security.

The broader impact extends to the global AI competitive landscape. The EU has swiftly used this event to advance its technology sovereignty agenda, emphasizing the urgency of reducing dependence on U.S. and Chinese technology. Other nations may follow the U.S. example, imposing stricter controls on their own AI companies in the name of national security, severely hindering open collaboration in global AI safety research.

Multiple Perspectives

Government Position: The White House emphasizes national security priority, arguing that models with critical vulnerabilities in the hands of malicious actors could have catastrophic consequences. Sacks stated Amodei’s refusal to cooperate was the direct trigger.

Amazon’s Role: As Anthropic’s largest investor and competitor, Amazon is both the provider of security research and a potential beneficiary. Its motivations raise industry questions about “regulatory capture.”

Anthropic’s Response: The company maintains the discovered vulnerabilities were “relatively simple” and present in other models equally, suggesting the government overreacted. But facing a 90-minute ultimatum, the company had virtually no negotiating room.

Industry Concerns: Multiple tech leaders warn such emergency controls could set a dangerous precedent, allowing the government to intervene in corporate product decisions at any time in the name of security, seriously damaging the innovation environment.

Editor: GoodInfo Global News Team