In a surprising development, Trump administration officials, including Treasury Secretary Scott Bessent and Federal Reserve Chair Jerome Powell, have reportedly encouraged major bank executives to test Anthropic’s new Mythos model. The directive, aimed at identifying potential vulnerabilities, saw top financial institutions like Goldman Sachs, Citigroup, Bank of America, and Morgan Stanley joining JPMorgan Chase in evaluating the capabilities of the Anthropic Mythos model.

Anthropic recently unveiled the Mythos model, noting its remarkable ability to detect security flaws, even though it wasn’t specifically trained for cybersecurity. This proficiency led Anthropic to initially limit access, citing concerns about its power. The situation is particularly intriguing given Anthropic’s ongoing legal dispute with the Trump administration over its designation as a supply-chain risk.

Despite this conflict, the encouragement for banks to test the Anthropic Mythos model suggests a complex interplay between government policy, national security, and the rapid advancement of AI technology. The financial sector’s embrace of this powerful AI tool, even amidst regulatory scrutiny, underscores its perceived value in bolstering defenses against sophisticated cyber threats.