FIS Launches AI Tool That Cuts Anti-Money Laundering Investigation Time by Over 90%

FIS Releases First-of-Its-Kind AI Tool for Anti-Money Laundering Investigations

FIS has announced the commercial availability to U.S.-based banks of its AML Investigations Agent, a first-of-its-kind AI tool developed in collaboration with Anthropic and design partner Amalgamated Bank1

1

2

2

. Originally announced in May 2026 as the Financial Crimes AI Agent, the solution is designed to streamline anti-money laundering investigations while keeping analysts firmly in control of every decision1

1

. The AI-driven agent is now being deployed at Amalgamated Bank, where early projections suggest it will dramatically reduce the time investigators spend on routine tasks2

2

.

Projected to Reduce Alert Review Time by Over 90%

Based on Amalgamated Bank's analysis of its investigation workflows combined with the FIS AML Investigations Agent's expected abilities, the agent is projected to reduce alert review time significantly across multiple investigation stages2

2

. Level 1 alert review time is expected to drop by more than 90%, while Level 2 case investigation time could decrease by approximately 98%2

2

. SAR narrative preparation time is projected to fall by more than 80%, freeing investigators to redirect capacity from data assembly to complex cases that require greater judgment2

2

. Stan Szwalbenest, head of enterprise fraud and financial crimes at FIS, explained that investigations slow down not because analysts lack judgment, but because evidence is scattered across systems and every alert gets treated the same1

1

.

How the Agent Can Distinguish Between Normal and Suspicious Activities

The FIS AML Investigations Agent applies customer-specific context to every investigation rather than treating all alerts the same1

1

2

2

. It maps the customer's type, historical activity and expected behavior against the alert before assembling evidence, so the resulting summary is already calibrated to what is and is not normal for that customer1

1

. The agent can distinguish between normal and suspicious activities with contextual precision. A seasonal spike in cash deposits looks different from deposits deliberately kept just under federal reporting thresholds, a practice known as structuring1

1

2

2

. Similarly, a routine international transfer looks different from an unexplained cross-border wire to a high-risk jurisdiction1

1

2

2

. The agent identifies those distinctions and explains the rationale to the analyst1

1

.

Built on Responsible AI Framework with Complete Audit Trail

The agent is built on FIS's Trust Assured framework, which embeds governance, compliance and responsible AI into every FIS product2

2

. Every output links to the specific record, document or check image it came from2

2

. Every recommendation, the reasoning behind it and the analyst's final decision are logged in a complete audit trail, ensuring investigation outcomes are explainable and ready for internal audit or regulatory scrutiny2

2

. Each institution's risk logic, thresholds and investigation procedures remain isolated through the FIS Control Plane, FIS's governance layer for managing agent access, controls and institution-specific configurations2

2

. This design ensures the agent augments, not replaces, the monitoring platforms and case management systems a bank already runs2

2

.

Integration with Core Banking Platforms and Claude Models

The FIS AML Investigations Agent natively integrates with FIS Horizon and FIS Systematics, two of FIS's core banking platforms, with FIS IBS planned next on the product roadmap2

2

. FIS developed and owns the agent, built with support from an Anthropic team that embedded with FIS product and compliance teams, drawing on FIS's financial-institution relationships, regulatory expertise and transaction-data infrastructure2

2

. The agent runs in FIS's private cloud and uses Anthropic's Claude models2

2

. The collaboration also produced the evaluation frameworks and engineering practices FIS will use to build and scale additional agents on the same governed platform2

2

. This matters because FIS already operates the core banking, payments and data infrastructure that financial crime investigations depend on, giving the agent access to comprehensive transaction data while maintaining embedded controls1

1

2

2

. Watch for how this deployment at Amalgamated Bank performs in real-world conditions and whether other U.S.-based banks rapidly adopt the technology to address growing anti-money laundering compliance burdens while improving fraud detection capabilities.