Financial technology company FIS has made an artificial intelligence agent designed to investigate potential money laundering commercially available to all U.S.-based banks, expanding access to a tool developed with Anthropic as financial institutions seek to automate costly compliance work.
The agent, developed with support from Anthropic and design partner Amalgamated Bank, is intended to reduce the manual work involved in reviewing suspicious transactions, collecting supporting evidence and preparing regulatory reports. Amalgamated Bank is deploying the system, FIS said.
The launch follows FIS’s initial announcement of its collaboration with Anthropic in May 2026. The company said the product is available to U.S. banks regardless of whether they already use FIS services.
FIS estimates the technology could cut the time required for initial anti-money laundering alert reviews by more than 90% and reduce second-level case investigation time by about 98%. Time spent preparing suspicious activity report narratives could fall by more than 80%.
The figures are projections based on Amalgamated Bank’s existing workflows and the agent’s expected capabilities, rather than measured results from its deployment.
The potential savings target a significant cost centre for banks. U.S. financial institutions spend an estimated $35 billion to $40 billion annually on anti-money laundering operations, according to 2023 estimates from LexisNexis cited by FIS.
Financial institutions face pressure to identify illicit transactions while managing large volumes of alerts generated by transaction-monitoring systems. Many alerts prove legitimate, requiring compliance teams to investigate activity that does not ultimately warrant further action.
Datos Insights estimated in 2023 that false-positive rates in the industry ranged from 90% to 95%, according to figures cited by FIS. Research by AnChain.AI in 2025 found that analysts spent 40% to 60% of investigation time gathering information before beginning their analysis, FIS said.
FIS said its agent evaluates alerts against a customer’s profile, transaction history and expected behaviour before gathering supporting evidence. The system is designed to help investigators distinguish routine activity from potentially suspicious transactions, including cash deposits deliberately kept below federal reporting thresholds.
Each finding is linked to its source record, document or cheque image. The system also records its recommendations, supporting reasoning and the final decisions made by analysts, creating an audit trail for internal reviews and regulatory examinations.
“Investigations don’t slow down because analysts lack judgment. They slow down because evidence is scattered across systems and every alert gets treated the same,” said Stan Szwalbenest, head of enterprise fraud and financial crimes at FIS.
Banks retain control over decisions
The agent operates alongside existing transaction-monitoring and case-management systems rather than replacing them. It integrates with FIS Horizon and FIS Systematics, while support for FIS IBS is planned.
FIS developed and owns the agent, which runs in its private cloud using Anthropic’s Claude models. An Anthropic team worked with FIS product and compliance teams during development, helping establish evaluation frameworks and engineering practices for future AI agents.
The system is built on FIS’s Trust Assured framework, which incorporates governance, compliance and responsible AI controls. Banks can retain their own risk thresholds and investigation procedures through the FIS Control Plane, which manages agent access and institution-specific configurations.
Human analysts retain responsibility for final decisions, including whether an alert warrants further investigation and whether a suspicious activity report should be filed.
In the United States, banks are required under the Bank Secrecy Act framework to monitor for suspicious activity and file reports with the Financial Crimes Enforcement Network when applicable legal thresholds are met. Automation can assist with collecting evidence and drafting reports, but does not remove banks’ compliance obligations.
The expansion of AI agents into financial crime compliance comes as banks and technology providers explore ways to automate tasks that require processing large volumes of transactions, customer records and supporting documents.
The potential benefits include faster investigations and more consistent documentation. However, banks must also assess whether AI-generated recommendations are accurate, explainable and sufficiently supported by evidence, particularly when decisions could affect customers or trigger regulatory reporting.
FIS has not disclosed the financial terms of its agreement with Anthropic, the agent’s pricing or the expected revenue contribution from the product. It also has not provided independently measured results from Amalgamated Bank’s deployment.
The commercial rollout is currently limited to banks based in the United States.
Stay ahead of the Stories shaping our world. Subscribe to Impact Newswire and join our
WhatsApp Channel for updates on global tech, business, and innovation—all in one place.
Dive deeper into the future with the Cause Effect 4.0 Podcast, where we explore the ideas, trends, and technologies driving the global AI conversation.
Got a story to share? Contact Us to reach a global audience with Impact Newswire.
Impact Newswire shares and distributes quality information on transactions that are likely or perceived to create greater impact.
Discover more from Impact Newswire
Subscribe to get the latest posts sent to your email.
