FIS and Anthropic Bring Agentic AI to Banking, Starting With AML Investigations

FIS is moving deeper into agentic AI for banking through a new collaboration with Anthropic, beginning with a Financial Crimes AI Agent designed for anti-money laundering investigations.

The financial technology company said the new agent will help banks compress AML alert and case investigations from hours or days to minutes by automatically assembling evidence, reviewing activity against known typologies, and surfacing higher-risk cases for investigator review. BMO and Amalgamated Bank are among the first institutions in development with the agent, while broader availability for FIS financial institution clients is planned for the second half of 2026.

Why FIS Is Starting With Financial Crimes

Financial crime compliance remains one of the most operationally heavy areas in banking. According to the announcement, the UN estimates that $2 trillion in illicit funds moves through the global financial system each year, while U.S. financial institutions spend an estimated $35 billion to $40 billion annually on AML operations.

A major challenge is that investigators often spend significant time gathering evidence from disconnected systems before they can begin meaningful analysis. FIS wants the Financial Crimes AI Agent to reduce that manual work by assembling case evidence automatically across bank systems through secure connections.

The agent is expected to evaluate activity against known financial crime typologies, reduce false positives, and improve the quality of investigative and Suspicious Activity Report narratives. FIS said investigators will remain in control of final decisions, with the agent used to support case preparation, prioritization, and review.

How Anthropic Fits Into the Strategy

Anthropic’s Applied AI team and forward-deployed engineers are embedded with FIS to co-design the Financial Crimes AI Agent. The work combines Claude’s reasoning capabilities with FIS’s banking data, compliance infrastructure, and regulatory expertise.

The collaboration is also designed to transfer knowledge to FIS so the company can build and scale additional AI agents independently over time. FIS said the Financial Crimes AI Agent is the first proof point for a broader roadmap of agentic AI tools built for bank-grade operations.

Jonathan Pelosi, Head of Financial Services at Anthropic, said FIS chose Claude because the model needed to reason through complex investigations, explain its work, and operate safely inside regulated workflows. He also said every conclusion reached by the agent links back to source data, while every decision stays with the investigator.

Governance and Auditability Are Central to the Model

For banks, the more important part of the announcement may not be the AI agent itself, but the governed environment around it.

FIS said client data will remain within FIS-controlled infrastructure at all times. The company is positioning itself as the data platform, governance layer, deployment infrastructure, and client relationship layer for regulated AI use in banking.

That matters because banks cannot treat AI deployment like a simple software add-on. In compliance and financial crime workflows, decisions must be explainable, auditable, and tied back to source data. FIS said every agent conclusion will be traceable and every decision will be recorded within its platform.

For financial institutions already using FIS systems, the agent is designed to access transaction, payment, deposit, credit, and customer activity data within the FIS environment. For banks using non-FIS core systems, the agent will connect through open integration standards.

What Comes Next

Financial crimes is only the starting point. FIS said its broader agent roadmap includes credit decisioning, deposit retention, customer onboarding, and fraud prevention.

The larger direction is clear: FIS is not simply adding AI assistance to banking workflows. It is building a governed agent platform where AI can act inside regulated environments while remaining traceable, controlled, and auditable.

For banks, this could become an important model for AI adoption. The opportunity is not just faster case handling. It is the possibility of reducing manual operational load while keeping compliance teams in control of high-risk decisions.

If FIS and Anthropic can deliver the promised investigation speed, governance, and auditability at scale, the Financial Crimes AI Agent could become an early example of how agentic AI enters core banking operations without bypassing regulatory discipline.

Key Source / Reference

Official source: Business Wire release — “FIS Brings Agentic AI to Banking with Anthropic, Starting with Financial Crimes.”

FAQ Section

What did FIS announce with Anthropic?

FIS announced that it is working with Anthropic to bring agentic AI to banking, starting with a Financial Crimes AI Agent for AML investigations.

What is the Financial Crimes AI Agent designed to do?

The agent is designed to assemble evidence across bank systems, evaluate activity against known typologies, reduce false positives, and help investigators review high-risk AML cases faster.

Which banks are first in development with the agent?

BMO and Amalgamated Bank are among the first institutions in development with the Financial Crimes AI Agent.

When will the FIS Financial Crimes AI Agent be broadly available?

FIS said broader availability for financial institution clients is planned for the second half of 2026.

What other banking areas could FIS target with AI agents?

FIS said its roadmap includes credit decisioning, deposit retention, customer onboarding, and fraud prevention.