Anthropic suspends all access to Mythos model after US government bans foreign nationals use

Anthropic Halts Mythos Access Amid US National Security Concerns

Anthropic suspends all access to Mythos – Following a directive from the US government, Anthropic has suspended access to its Mythos model, marking a significant shift in the company’s operations. The decision, announced in a statement on June 13, comes as part of broader efforts by the Trump administration to tighten controls on AI technologies. While the suspension initially targeted foreign nationals, it now applies to all users, including internal teams. Anthropic emphasized that the move was prompted by concerns over cybersecurity risks, though the exact rationale behind the government’s mandate remains unclear.

Expanded Restrictions and Industry Impact

The ban on Mythos access represents one of the most sweeping actions taken by the US government to date, affecting not just foreign users but also domestic ones. Anthropic’s statement noted that the restriction was based on the potential for the model to be exploited in ways threatening national security. This move aligns with the administration’s push to ensure AI systems are aligned with strategic interests, with analysts suggesting it may signal a broader trend of regulatory scrutiny on advanced models.

“The government’s directive has forced us to re-evaluate our model’s accessibility, even though the vulnerabilities identified are not unique to Mythos.”

Despite the suspension, Anthropic maintained that the security concerns were not enough to justify a full ban. The company highlighted that similar weaknesses exist in other publicly available systems, arguing the measure could hinder innovation. This stance reflects a growing tension between regulatory oversight and the need for widespread AI adoption in critical sectors.

Mythos and the Path to National Security

Mythos, Anthropic’s latest AI model, was initially designed for secure, high-stakes applications such as protecting critical infrastructure. Its advanced capabilities, including the ability to exploit software vulnerabilities, made it a focal point for government attention. The model’s transition from restricted use to broader availability through the Fable 5 version has sparked debates about balancing security with accessibility in AI development.

While the ban on foreign access was part of the original directive, the scope has since expanded to include all users. This has raised questions about the practicality of the restriction, as Anthropic must now secure licenses for domestic transfers as well. The Commerce Department, which issued the order, has not provided detailed justification, leaving industry stakeholders to speculate on the underlying motivations.

Government’s Regulatory Strategy

The US government’s action against Mythos is part of a larger strategy to monitor AI technologies. This includes the recent executive order mandating that companies share advanced models with federal agencies before wider distribution. Anthropic’s suspension underscores the administration’s focus on preventing AI from being used in ways that could compromise national security, particularly in military and financial contexts.

Mythos’s release had coincided with the administration’s push for greater oversight of AI systems. While the foreign nationals’ access ban was a proposed element of the executive order, it was implemented more broadly than initially outlined. The move has been met with mixed reactions, with some viewing it as necessary for security and others criticizing it for stifling innovation.

Corporate Response and Legal Context

Anthropic has expressed cooperation with the government’s request to “red team” its models, acknowledging the importance of security assessments. However, the company argues the current restrictions are overly broad, impacting users who pose no direct threat. This aligns with Anthropic’s history of challenging regulatory decisions, including its successful legal battle against a previous designation as a supply chain risk.

The suspension of Mythos access highlights the evolving relationship between AI firms and government regulators. While the Trump administration has sought to impose stricter controls, Anthropic’s executives continue to engage with policymakers, demonstrating a balance between compliance and advocacy for the industry’s growth. The company remains optimistic about the long-term benefits of AI despite the immediate restrictions.