The U.S. Department of Commerce has ordered Anthropic to suspend its most advanced artificial intelligence models, Fable 5 and Mythos 5, citing cybersecurity risks and the potential to bypass internal safety controls. The directive took effect June 13 and applies globally, forcing the company to take both systems offline for domestic and international users.
The move followed a warning from Amazon chief executive Andy Jassy, who told the White House that safeguards in Fable 5 could be circumvented to generate information useful for network attacks. The issue quickly escalated to senior national security officials, including Treasury Secretary Scott Bessent.
national security concerns override company assurances
Anthropic chief executive Dario Amodei argued during multiple calls with U.S. officials that the issue was limited to a specific vulnerability rather than a systemic flaw. However, the administration, supported by an assessment from the National Security Agency, determined the risk justified immediate action.
According to officials, Anthropic declined to voluntarily withdraw the models before export controls were imposed. The company said it was given just 90 minutes to comply and was not provided with detailed information about the threat. The White House countered that the decision came only after extended discussions failed to produce agreement.
The action marks a significant shift in how Washington treats advanced AI, framing it as a matter of national security oversight rather than corporate responsibility.
amazon’s role draws attention
Amazon’s involvement has come under scrutiny as both a major backer and infrastructure provider for Anthropic. By late 2024, the company had invested $8 billion and supplied key cloud resources used to train the models.
At the same time, Amazon has explored potential investments of up to $50 billion in OpenAI, placing it between competing AI developers and raising questions about its influence across the sector.
rising tension between Anthropic and Washington
The suspension follows earlier friction. In March, the Pentagon labeled Anthropic a supply chain risk after the company refused to support certain government uses of its systems, including domestic surveillance and autonomous weapons.
Amodei’s previous comments comparing advanced AI risks to nuclear weapons also drew attention. Officials reportedly viewed those remarks, combined with the refusal to pause operations during a vulnerability review, as a compliance concern rather than a purely technical issue.
market reaction highlights centralized risk
The shutdown underscores how quickly government intervention can disrupt leading AI platforms. Anthropic’s pre-IPO perpetual contract on the Hyperliquid exchange fell about 3.7% to roughly $1,627 following the news, signaling a reassessment among traders of the company’s valuation outlook.
At the same time, capital flowed into decentralized alternatives. In the week after the suspension, an estimated $2.87 billion moved into AI-focused crypto tokens, with most major assets in the category posting gains. Render, an AI-linked cryptocurrency, rose nearly 4%, reflecting a shift toward infrastructure less exposed to centralized control.
new precedent for AI regulation
Washington’s decision represents the first major use of a recent executive framework that encourages AI developers to share advanced models with federal agencies for review before release. The Anthropic case demonstrates the government’s willingness to enforce compliance and restrict access when risks are identified.
For companies integrating advanced AI systems, the episode highlights a growing operational risk: access to critical tools can be revoked abruptly. The incident also raises broader concerns about vendor dependency, especially when infrastructure providers have direct lines to policymakers.
The shutdown leaves competitors such as OpenAI in a stronger position while forcing Anthropic to navigate regulatory scrutiny that now treats its core technology as a strategic asset under national security control.
Concerned about AI security and regulation? Explore how robust risk controls protect digital platforms and users.
Disclaimer: The content on this page is provided for general informational purposes only and does not represent the views or financial advice of Toobit. We make no guarantees regarding the accuracy or completeness of this information and shall not be held liable for any errors, omissions, or outcomes resulting from its use. Investing in digital assets involves risk; users should independently evaluate their financial situation and the risks involved. For further details, please consult our Terms of Service and Risk Disclosure.
