FIS Partners with Anthropic to Build AI Agent for Banking

What happened

FIS announced a partnership with Anthropic to develop a Financial Crimes AI Agent designed to automate anti-money laundering (AML) investigations for financial institutions. The agent uses Anthropic's Claude AI model to automatically gather evidence across a bank's core systems, evaluate suspicious activity against known patterns of illicit behavior, and surface high-risk cases for investigator review. According to FIS, the system can reduce investigation time from hours or days to minutes. BMO and Amalgamated Bank are among the first institutions developing deployments of the agent. Anthropic's Applied AI team and forward-deployed engineers are working directly with FIS to co-design the system. FIS plans broader commercial availability in the second half of 2026. The company stated that customer data remains inside FIS-controlled infrastructure and all AI-generated decisions remain traceable and auditable. FIS described the initiative as part of a broader agent-first banking strategy, with future agents planned for credit decisioning, deposit retention, customer onboarding, and fraud prevention.

Why it matters

FIS positions itself as a trusted intermediary layer between banks and AI, managing data governance and regulatory compliance while providing access to advanced AI capabilities. This approach addresses a critical pain point in banking: U.S. financial institutions currently spend an estimated $35 billion to $40 billion annually on AML operations, with investigators spending most of their time manually gathering information across disconnected systems. By automating evidence gathering and case prioritization, FIS aims to reduce investigation costs and allow human investigators to focus on higher-risk decisions. The partnership leverages FIS' role as a system of record for transactions, deposits, credit activity, payments, and customer information across thousands of financial institutions globally. The embedded collaboration model with Anthropic is designed to enable FIS to eventually build and scale additional agents independently, giving the company a degree of technical independence that standard API integrations typically do not provide. This matters for regulatory compliance, as banks must demonstrate to examiners that personnel understand how AI models reach their conclusions.

Bigger picture

The partnership reflects broader efforts across the financial services industry to deploy AI in regulated workflows while maintaining compliance and auditability. Financial crime compliance costs in the United States and Canada exceed $61 billion annually, and the United Nations estimates that up to $2 trillion in illicit funds moves through the global financial system each year. Traditional rules-based AML systems often produce false positive rates above 99%, creating significant operational burden for banks. AI agents represent an emerging category of automation in banking, but regulatory bodies have made clear through guidance and enforcement actions that automation does not reduce accountability. Any bank deploying AI agents must demonstrate clear reasoning chains for decisions and maintain human oversight. FIS' approach of positioning itself as a governed platform through which banks access AI, rather than a standalone tool vendor, addresses these regulatory requirements. The company's roadmap to expand beyond financial crimes into credit decisioning, fraud prevention, and customer onboarding suggests AI agents could reshape multiple areas of banking operations in the coming years.

What to watch

Monitor the pilot deployments at BMO and Amalgamated Bank for operational results and regulatory feedback before the planned broader launch in the second half of 2026. Watch whether FIS successfully transitions from co-developing agents with Anthropic to building and scaling agents independently, as the company claims. Observe regulatory responses as banks deploy AI agents in AML workflows, particularly whether examiners accept the traceability and auditability frameworks FIS has designed. Track whether FIS delivers on its product roadmap to expand AI agents into credit decisioning, deposit retention, customer onboarding, and fraud prevention. Pay attention to competitive responses from other banking technology providers and whether FIS' positioning as a governed intermediary layer resonates with financial institutions.

This article was generated by Quantli AI using publicly available news sources.