Powell and Bessent Convene U.S. Bank CEOs Amid Concerns Over Anthropic’s Mythos AI Cyber Threat
April 10, 2026 — In a rare and urgent meeting this week, Federal Reserve Chair Jerome Powell and Treasury Secretary Scott Bessent gathered the CEOs of major U.S. banks to address growing cybersecurity concerns related to Anthropic’s recently launched artificial intelligence model, Mythos.
The invitation-only session took place in Washington, D.C., coinciding with a Financial Services Forum event where many banking leaders were already present. The meeting, held at the Treasury Department on Tuesday evening, brought together top executives including Bank of America’s Brian Moynihan, Citigroup’s Jane Fraser, Goldman Sachs’ David Solomon, Morgan Stanley’s Ted Pick, and Wells Fargo’s Charlie Scharf. JPMorgan Chase CEO Jamie Dimon was notably absent.
Addressing the Cybersecurity Risks of Mythos AI
Anthropic, an AI-focused startup known for pushing the envelope on advanced machine learning models, rolled out its new Claude Mythos Preview earlier this week—but deliberately limited its deployment due to fears that malicious actors could exploit the technology. The capabilities of Mythos extend into both offensive and defensive cyber applications, prompting warnings about potential abuse by hackers.
Prior to the release, Anthropic had engaged in briefings with senior U.S. government officials to discuss the model’s implications, a company spokesperson confirmed to CNBC. Dialogue has involved agencies such as the Cybersecurity and Infrastructure Security Agency (CISA) and the Center for AI Standards and Innovation. In a public statement, Anthropic CEO Dario Amodei emphasized the fine balance between risk and opportunity: getting the technology right could usher in a more secure digital environment, whereas failure to manage its use properly “carries obvious dangers.”
Government and Industry’s Heightened Vigilance
The high-profile gathering underscores the Trump administration’s growing alarm over AI’s potential to disrupt the financial system and critical infrastructure. Powell and Bessent’s involvement signals that the threat posed by AI-enabled cyberattacks is on the radar of the nation’s highest monetary and economic policymakers.
This concern arises amid Anthropic’s ongoing conflict with the Department of Defense (DoD), which last month designated the company as a supply chain security risk. President Donald Trump and Defense Secretary Pete Hegseth have publicly criticized Anthropic for setting restrictions on how its AI can be used in military contexts. Despite the official blacklisting, evidence suggests the DoD has continued using Anthropic’s Claude AI in operations related to the Iran conflict.
Anthropic’s legal challenges seeking to overturn the DoD’s designation remain unresolved. While a federal appeals court recently denied a temporary block of the blacklisting, an earlier preliminary injunction allows the company to provide services to other government agencies as far as legal battles proceed.
Broader Industry Implications and Partnerships
JPMorgan Chase is among Anthropic’s initial partners for Project Glasswing, a cybersecurity initiative involving heavyweights like Apple, Google, Microsoft, and Nvidia. This collaboration aims to harness AI’s capabilities to strengthen cyber defenses without amplifying vulnerabilities.
However, concerns have mounted following reports that hackers have already exploited Anthropic’s AI in real-world attacks. Last November, Anthropic disclosed that a Chinese hacking group leveraged its Claude model to automate intrusions into government and corporate networks, intensifying scrutiny on the company’s technology.
Market and Industry Reaction
The unveiling of Mythos and Anthropic’s disclosures caused an immediate ripple through cyber stock markets, which suffered declines after a leaked draft blog post highlighted the AI model’s dual-use potential for both enhancing and breaching cybersecurity.
As regulators, government leaders, and industry stakeholders convene to navigate the challenges posed by AI-driven cyber threats, the Anthropic Mythos episode exemplifies the delicate balance between innovation and security in the digital age.
For more updates on AI technology and cybersecurity developments, follow CNBC Tech and subscribe to CNBC PRO.