Federal Agencies Quietly Test Claude Mythos Despite Trump Ban

Federal agencies are quietly testing Anthropic's Claude Mythos AI model despite restrictions from the Trump administration, according to sources familiar with the matter. At least two US federal agencies and three congressional committees have reached out to Anthropic to evaluate the advanced AI system, effectively bypassing the administration's ban on certain AI technologies.

The Commerce Department's Center for AI Standards and Innovation, which is tasked with evaluating both US and foreign AI models, appears to be among the entities conducting these tests. This represents a significant development in the ongoing tension between federal agencies seeking to leverage cutting-edge AI capabilities and the administration's restrictive policies on AI deployment.

The Testing Circumvention

The testing of Claude Mythos by federal entities represents a notable workaround of the Trump administration's AI restrictions. While the administration has implemented various limitations on AI adoption and deployment across federal agencies, these entities appear to be finding ways to evaluate advanced models like Mythos for potential use cases.

This quiet testing effort suggests that federal agencies recognize the strategic importance of staying current with AI capabilities, even when faced with administrative barriers. The involvement of congressional committees in the testing process also indicates that there is significant interest in understanding the capabilities and limitations of advanced AI systems at the legislative level.

Treasury Department's Interest

Separately, the US Treasury Department's technology team is seeking to gain access to Anthropic's Mythos AI model as soon as this week. Treasury CIO Sam Corcos has directed the department's cybersecurity team to prepare for AI threats, recognizing the growing importance of AI in both offensive and defensive cybersecurity operations.

The Treasury's interest in Mythos appears to be focused on using the AI system to hunt for vulnerabilities and potential threats within the department's vast financial systems. This represents a practical application of advanced AI for cybersecurity purposes, particularly in protecting critical financial infrastructure.

Broader Context of AI Adoption

The federal testing of Claude Mythos occurs against a backdrop of rapid AI advancement and increasing adoption across both public and private sectors. Anthropic has been making significant strides with its Claude models, including recent releases like Claude Opus 4.7 and various tools for website and presentation design.

However, the company has faced criticism from some users who accuse it of degrading model performance to manage capacity. Anthropic employees have publicly denied these allegations, but the controversy highlights the challenges of scaling AI systems to meet growing demand while maintaining consistent performance.

Market Implications

Anthropic's growing prominence in the AI landscape is reflected in its soaring valuation. The company has recently received multiple offers from venture capital firms valuing it at as much as $800 billion, up from $380 billion in February. This dramatic increase in valuation underscores the market's confidence in Anthropic's technology and its potential for future growth.

The company is also preparing to release Claude Opus 4.7, along with new AI-powered tools for designing websites and presentations. These developments suggest that Anthropic is continuing to push the boundaries of what's possible with large language models while expanding into new application areas.

Policy Tensions

The federal testing of Claude Mythos highlights the ongoing tension between technological advancement and regulatory oversight. While the Trump administration has implemented various restrictions on AI adoption, federal agencies and congressional committees appear to be pursuing their own evaluations of advanced AI systems.

This situation reflects a broader challenge facing governments worldwide as they grapple with how to regulate rapidly evolving AI technologies while ensuring that public sector entities can leverage these tools effectively. The quiet testing of Mythos by federal agencies suggests that practical needs may sometimes override administrative restrictions in the pursuit of technological capability.

Looking Forward

As federal agencies continue to evaluate Claude Mythos and other advanced AI systems, the tension between innovation and regulation is likely to persist. The outcome of these evaluations could influence future policy decisions regarding AI adoption in the public sector.

The Treasury Department's focus on using AI for cybersecurity purposes also points to a growing recognition of AI's dual-use nature - as both a potential threat and a powerful defensive tool. This dual nature will likely shape how federal agencies approach AI adoption in the coming years.

For Anthropic, the federal interest in Claude Mythos represents both an opportunity and a challenge. While government adoption could provide significant validation and revenue opportunities, it also comes with increased scrutiny and the need to navigate complex regulatory environments.

The quiet testing of Claude Mythos by federal agencies, despite administrative restrictions, underscores the powerful pull of advanced AI technology and the challenges of controlling its adoption in an era of rapid technological change.