Trump Administration Blocks International Access to Anthropic AI Models Following Amazon's Security Alert

TLDR

Table of Contents

Andy Jassy, Amazon’s CEO, alerted Trump administration officials after internal research revealed Anthropic’s Fable 5 system could facilitate malicious cyber operations
White House officials demanded Anthropic address the security flaws or limit model availability; President Trump subsequently approved complete international access restrictions
To meet export control requirements, Anthropic disabled both Fable 5 and Mythos systems for all users globally
The company characterized the security gaps as “relatively basic” and noted comparable features are available in other publicly accessible AI platforms
The restrictions threaten Anthropic’s planned public offering while potentially strengthening competitors like OpenAI

Andy Jassy, serving as Amazon’s chief executive, engaged in direct communications with U.S. government representatives, including Treasury Secretary Scott Bessent, regarding potential security vulnerabilities discovered in Anthropic’s Fable 5 system. Internal Amazon security teams had successfully employed targeted prompts to extract information from the model that could facilitate malicious cyber operations.

The US government, citing national security authorities, has issued an export control directive to suspend all access to Fable 5 and Mythos 5 by any foreign national, whether inside or outside the United States, including foreign national Anthropic employees.

The net effect of…

— Anthropic (@AnthropicAI) June 13, 2026

Such sensitive data should have been prevented from disclosure by the model’s built-in safety mechanisms. These discoveries were promptly communicated to White House personnel and national security advisors, triggering a series of high-level deliberations on appropriate countermeasures.

Government representatives presented Anthropic with an ultimatum: remediate the security weaknesses or suspend the model’s operations. During Friday consultations between Anthropic’s CEO Dario Amodei and administration representatives, certain officials perceived the company as resistant to collaborating with government cybersecurity specialists on developing solutions.

Amazon.com, Inc., AMZN

Administration officials determined that implementing foreign access restrictions represented the most effective approach to mitigating potential risks. President Trump authorized the decision, though he privately expressed concerns about potential negative impacts on artificial intelligence advancement.

Anthropic Deactivates Premier AI Systems

In response to the government directive, Anthropic terminated access to both Fable 5 and its Mythos platform across all user categories — extending beyond international users — to guarantee full adherence to newly implemented export regulations.

The organization described the identified vulnerabilities as comparatively unsophisticated. Anthropic emphasized that similar information-retrieval capabilities exist within other commercially available AI systems.

Cybersecurity expert Andrew Morris, who established GreyNoise Intelligence, examined Amazon’s research data. His assessment indicated that while Fable 5 demonstrated the ability to detect software vulnerabilities across at least four applications, no evidence suggested researchers successfully accessed the more critical functionality of transforming identified bugs into executable attack tools.

Anthropic has consistently promoted safety protocols as a fundamental principle. The organization previously postponed wider distribution of Mythos following White House guidance and maintains regular coordination with a specialized government AI evaluation division prior to launching new systems.

What This Means for Anthropic

The development arrives at a particularly challenging moment for Anthropic. The organization is currently preparing for a possible initial public offering potentially scheduled for autumn this year.

With its premier AI systems unavailable, existing clients may migrate to alternative providers. OpenAI maintains its own cybersecurity-oriented model and has sustained ongoing communications with the Trump administration.

Friction between Anthropic and White House officials predates this incident. The Pentagon had previously classified Anthropic as a security concern, a determination the company is challenging through two concurrent legal proceedings.

National Cyber Director Sean Cairncross and Commerce Secretary Howard Lutnick participated in the deliberations that resulted in the access prohibition. The Commerce Department maintains regulatory authority over export restrictions affecting sensitive technologies.

Commerce’s prohibition currently blocks foreign governmental entities, corporations, and private individuals from utilizing Fable and Mythos systems. A significant portion of Anthropic’s research personnel were born outside the United States, which the company indicated effectively prevents their participation in model development activities.

White House AI adviser David Sacks characterized the restriction as implemented “reluctantly” and conveyed optimism that Anthropic would address the concerns, enabling the models’ restoration to unrestricted availability.

✨ Limited Time Offer