An AI agent is about to police millions of bank accounts for financial crimes, and it runs on Claude.
Anthropic and Fidelity National Information Services (FIS), the fintech giant that processes nearly 12% of the global economy, announced a partnership Monday to build agentic AI systems for banks. The first product: a Financial Crimes AI Agent that compresses anti-money-laundering investigations from hours to minutes. The math explains why banks are interested. U.S.
financial institutions spend $35-40 billion annually on AML compliance, yet investigators waste most of their time manually stitching together evidence from disconnected systems before any real analysis starts. The Financial Crimes AI Agent automates that assembly work across a bank's core systems, evaluates activity against known typologies, and surfaces only the highest-risk cases for human review.
FIS CEO Stephanie Ferris confirmed the agent targets drug traffickers, terrorists, and other criminals using the financial system. Anthropic's Applied AI team and forward-deployed engineers are embedded at FIS to co-design the tool, combining Claude's reasoning with FIS's proprietary banking data and regulatory infrastructure.
"The launch of AI agents marks a new era in banking," Ferris said in the release, adding that financial institutions have an opportunity to become "agent-first."
Bank of Montreal and Amalgamated Bank will be the first to deploy the agent, with broader availability planned for the second half of 2026. FIS shares jumped about 7% in after-hours trading after The Wall Street Journal reported the partnership. The human element remains intact. Ferris emphasized that investigators will still make final decisions on whether to pursue cases. The agent handles the grunt work, gathering transaction data, account information, and evidence scattered across multiple systems, but humans keep the authority.
This isn't a one-off experiment. FIS plans to build additional agents for credit decisioning, deposit retention, customer onboarding, and fraud prevention, all delivered through a single governed platform.
Anthropic's head of industries, Jonathan Jager-Hyman, confirmed the company's engineers are already embedded within FIS teams. The EU is also watching closely. Bloomberg reported that European Economy Minister Valdis Dombrovskis confirmed the bloc is in talks with Anthropic about getting banks and companies tested for vulnerabilities uncovered by the company's Mythos AI model.
"Indeed there are contacts with Anthropic," Dombrovskis said Monday night after a Eurogroup meeting in Brussels. The UN estimates $2 trillion in illicit funds flows through the global financial system annually. Banks already spend billions on AML compliance. The bet from Anthropic and FIS is that AI agents can finally move the needle, not by replacing compliance teams, but by letting them focus on actual crime instead of paperwork.
