The Federal Reserve was without access to Anthropic's Mythos AI model for several months, even as major banks moved quickly to integrate the technology into their cybersecurity defenses.
The central bank had raised alarms about the model's capabilities in April during an extraordinary meeting with top bank CEOs, yet it did not gain access to the tool until mid-July.
This delay underscores a significant gap in the Fed's ability to monitor and regulate the rapid adoption of generative AI within the financial system.
While institutions like JPMorgan Chase and Amazon were racing to patch vulnerabilities and leverage the new model, the regulator remained in the dark, limiting its capacity to assess systemic risks in real time.
The revelation comes as cybersecurity firms see their recent rally falter, despite Anthropic positioning Mythos as a catalyst for renewed demand in the sector.
The disconnect between private-sector adoption and regulatory oversight raises questions about the Fed's preparedness for the next wave of AI-driven financial innovations.