Anthropic is moving to make its Mythos artificial intelligence model available to European and United Kingdom banks, according to people familiar with the matter cited in recent reporting. The company intends to broaden access beyond the handful of U.S. institutions already using the technology, but officials said the expansion will proceed only after security checks are completed.
Cybersecurity specialists have flagged Mythos as presenting significant challenges to banking institutions and their legacy technology frameworks. Those concerns were raised alongside regulatory and policymaker warnings at last week’s International Monetary Fund spring meeting in Washington, where participants discussed risks associated with emerging AI tools in the financial sector.
Sources say Anthropic has begun the process of offering Mythos to European banks and other organizations, with the rollout contingent on a series of safeguards intended to protect sensitive systems. One person familiar with the preparations indicated that access for European banks could be arranged within days. Another individual cautioned that the deployment may take days or weeks, reflecting remaining technical and security validation steps.
Bloomberg has separately reported that Anthropic plans to make Mythos available to financial institutions in the U.K. in the near term. Within the United States, a number of banks have already been provided with access to Mythos as other lenders work to evaluate and integrate the model.
Anthropic has publicly acknowledged that JPMorgan Chase (JPM.N), which participates in the Glasswing initiative, has access to the Mythos model. Bank of America (BAC.N) has also been involved with Glasswing from the outset and has been conducting internal tests of the Mythos technology, according to people familiar with the matter.
The company’s stated approach emphasizes completing checks and controls before broadening availability, reflecting ongoing industry caution about how advanced AI systems interact with incumbent banking technology and risk-management processes.
Summary
Anthropic plans to extend Mythos AI access to European and U.K. banks, subject to security checks; cybersecurity experts and policymakers have expressed concerns about the model's implications for banking systems, and a group of U.S. banks have already gained access.
Sectors affected: Banking, cybersecurity, financial technology.