WASHINGTON—U.S. banking regulators are delaying some cyber-related examinations of the nation’s largest banks as financial institutions and supervisors assess new risks tied to Anthropic Mythos artificial intelligence model, according to one new report.
Bloomberg reported the Federal Reserve and the Office of the Comptroller of the Currency are giving banks additional time to strengthen cyber defenses as regulators and lenders test the capabilities and risks associated with the AI technology. The report cited people familiar with the matter who requested anonymity because the discussions are private.
According to Bloomberg, Anthropic said last month it would limit access to Mythos after warning the model could potentially be used to facilitate cyberattacks. The company also launched an initiative known as Project Glasswing with select firms, including Apple and JPMorgan Chase, allowing them to test the technology against their own cybersecurity systems before broader deployment.
Bank, Government Meetings
As the CU Daily reported earlier, Treasury Secretary Scott Bessent and then-Federal Reserve Chair Jerome Powell met with Wall Street executives in April to discuss cyber threats posed by advanced AI models such as Mythos. Since then, regulators and banks have been working on ways to address vulnerabilities exposed by the technology, according to Bloomberg.
Inside major financial institutions, executives initially reacted with alarm to Mythos’ ability to rapidly analyze code and identify potential security weaknesses, Bloomberg reported. But after weeks of testing, some of the initial concern has shifted toward longer-term remediation efforts and system reviews.
Time to Understand the Tech
Bloomberg said the delay in some cyber examination schedules is intended to provide banks with more time to understand the technology while also helping regulators conduct stress tests of existing defenses. The OCC is also conducting its own testing of Mythos, according to Bloomberg’s sources.
The OCC declined comment to Bloomberg, while the Federal Reserve did not immediately comment.
Federal Reserve Vice Chair for Supervision Michelle Bowman said earlier this month that regulators would continue refining oversight of emerging technologies such as Mythos.
“Regulators will continue to focus on critical developments and communicating these risks to supervised institutions, as well as on refining our cybersecurity approach,” Bowman said, according to Bloomberg.
Special Teams Formed
Bloomberg also reported that several major banks, including Morgan Stanley and Goldman Sachs, have formed specialized teams focused on testing Mythos and coordinating with federal intelligence agencies and outside security vendors.
