Trump officials urge banks to test Mythos AI, an advanced model from Anthropic designed to detect cybersecurity vulnerabilities, despite the Department of Defense previously designating Anthropic as a supply-chain risk. This unprecedented move by the Trump administration signals a significant shift in how Wall Street is being encouraged to approach next-generation cyber threats, pushing for the adoption of cutting-edge artificial intelligence even amidst prior federal warnings.
Reports indicate that top economic officials, including Treasury Secretary Scott Bessent and Federal Reserve Chair Jerome Powell, have actively lobbied major US financial institutions to integrate Anthropic’s Mythos model into their cybersecurity protocols. White House national economic adviser Kevin Hassett further corroborated the administration’s proactive stance, emphasizing efforts to shield the public from escalating AI-driven risks. The encouragement comes at a critical juncture, as financial institutions face an increasingly sophisticated threat landscape.
The Mythos Model: A Double-Edged Sword?
The Mythos model, an unreleased internal Anthropic AI project code-named “Project Glasswing,” has demonstrated extraordinary capabilities. It can autonomously discover thousands of zero-day security vulnerabilities across major operating systems and web browsers. Anthropic itself has cautioned against its widespread public release without robust safeguards, acknowledging its inherent danger. The tension between its offensive potential and defensive application is a central theme in its current deployment.
“Mythos’s capabilities are too dangerous for public release without better safeguards.”
Major Wall Street players like JPMorgan Chase & Co., Goldman Sachs Group Inc., Citigroup Inc., Bank of America Corp., and Morgan Stanley are reportedly conducting internal tests, assessing Mythos’s capacity to scan faster than human teams and predict complex attack paths. This internal testing is crucial for understanding how such powerful AI can be safely integrated into critical financial infrastructure, especially given the model’s self-professed hacking prowess.
Conflicting Directives and National Security Concerns
The administration’s push to get banks to test Mythos AI stands in stark contrast to previous directives. In February and March 2026, the Pentagon, under Defense Secretary Pete Hegseth, labeled Anthropic a “supply chain risk to national security.” This designation effectively barred any contractor or partner doing business with the U.S. military from engaging commercially with Anthropic. The disagreement reportedly stemmed from Anthropic’s concerns about its AI being used for mass domestic surveillance and autonomous weapons, which the Pentagon deemed vital for national defense.
Adding to the complexity, conflicting court rulings have muddied the waters. A San Francisco judge initially ordered the removal of the supply-chain risk label, suggesting the DoD may have acted in bad faith. However, a US appeals court in Washington, D.C., later ruled that Anthropic had not met the requirements to temporarily lift the designation. This legal back-and-forth highlights the significant policy and ethical dilemmas surrounding advanced AI development and deployment, particularly when national security interests intersect with private sector innovation. For more related Tech news, visit our Tech news section.
Why Financial Institutions Need Advanced Cyber Defense
The primary driver behind the administration’s encouragement for banks to test Mythos AI is the escalating threat of financial cyberattacks. Traditional cybersecurity measures are struggling to keep pace with sophisticated, AI-driven threats. Mythos offers a potential leap forward, providing an AI-powered solution that can detect hidden financial cyber threats and prepare institutions for next-generation attacks with unparalleled speed and precision. The ability of Mythos to autonomously identify and exploit vulnerabilities, even if initially concerning, can be repurposed for robust defensive strategies, securing critical software and financial systems against future breaches.
The convergence of advanced AI capabilities, national security concerns, and the imperative to protect financial stability has placed Anthropic’s Mythos model at the center of a complex and evolving debate. As Wall Street banks continue their internal evaluations, the outcomes will undoubtedly shape future policies regarding AI adoption in critical infrastructure, balancing innovation with the inherent risks of powerful, autonomous technologies.
