TLDR
- Treasury Secretary Scott Bessent and Fed Chair Jerome Powell held an urgent meeting with Wall Street bank CEOs over Anthropic’s Mythos AI model
- Mythos can identify and exploit software vulnerabilities across every major operating system and browser
- CEOs from Citigroup, Morgan Stanley, Bank of America, Wells Fargo, and Goldman Sachs attended; JPMorgan’s Jamie Dimon could not join
- Access to Mythos is limited to around 40 tech companies including Microsoft and Google under “Project Glasswing”
- Crypto and DeFi experts have raised concerns about Mythos targeting zero-day vulnerabilities in decentralized infrastructure
U.S. Treasury Secretary Scott Bessent and Federal Reserve Chair Jerome Powell called an urgent meeting with top Wall Street bank CEOs this week to discuss cybersecurity risks tied to Anthropic’s new AI model, Mythos.
Anthropic Model Scare Sparks Urgent Bessent, Powell Warning to Bank CEOs
US Treasury Secretary Scott Bessent and Federal Reserve Chair Jerome Powell held an urgent meeting with Wall Street leaders to address cybersecurity risks linked to Anthropic’s Mythos AI model. Bank CEOs… pic.twitter.com/1ojYAJOztU— CN Wire (@Sino_Market) April 10, 2026
The meeting took place Tuesday in Washington. It was hosted by the Treasury Department and aimed at making sure banks understood the risks posed by Mythos and were taking steps to protect their systems.
CEOs from Citigroup, Morgan Stanley, Bank of America, Wells Fargo, and Goldman Sachs were present. JPMorgan CEO Jamie Dimon was unable to attend, according to one source familiar with the matter.
Invitations went out while many of the bank chiefs were already in Washington for separate meetings, making the gathering easier to arrange on short notice.
All five banks that attended are classified as systemically important financial institutions. That means disruptions to their operations could have consequences that ripple across the global financial system.
Mythos is an AI model built by Anthropic. It is designed to find and exploit weaknesses in software systems. Unlike consumer-facing AI tools, Mythos is focused on cybersecurity and software engineering tasks.
What Makes Mythos Different
Anthropic has said Mythos is capable of identifying and exploiting vulnerabilities across every major operating system and every major web browser.
The model can spot what are known as zero-day vulnerabilities — flaws in software that have not yet been publicly discovered or patched. It can then build working exploits from those findings.
Anthropic launched Mythos earlier this week but stopped short of a wide release. The company said it was limiting access due to the model’s potential to expose unknown cybersecurity weaknesses.
Access is currently restricted to around 40 technology companies, including Microsoft and Google, under a program called “Project Glasswing.”
Anthropic said it proactively briefed senior U.S. government officials and key industry stakeholders on Mythos’s capabilities before the release.
Concerns Extend to Crypto and DeFi
The risks are not limited to traditional banks. Experts in crypto and decentralized finance have flagged concerns that Mythos could be used to target vulnerabilities in DeFi infrastructure.
The worry is that the model can discover and exploit zero-day flaws in real-time and at low cost, which could threaten decentralized protocols that rely on smart contract security.
Anthropic is also currently in a legal dispute with the Pentagon. The U.S. Department of Defense has designated the company a supply-chain risk, a classification Anthropic is contesting in court.
Goldman Sachs, Wells Fargo, and the Federal Reserve all declined to comment. The Treasury Department, the attending banks, and Anthropic did not immediately respond to requests for comment.
