Fintech FIS and Anthropic are joining forces to build an AI agent for anti-money laundering in banking.
FIS began developing the solution after observing that the “manual burden of evidence assembly” was the primary challenges facing its clients in anti-money laundering (AML) processes, Matt Valente, senior vice president of corporate strategy at FIS, told FinAi News.
“Investigators were spending the majority of their time pulling data from disconnected systems before any real analysis could begin,” he said. “That’s a structural problem, and it’s exactly where an AI agent — one that can autonomously assemble a complete evidence package across a bank’s core systems at case open — delivers measurable, immediate value.”
BMO and Amalgamated Bank will be among the first to deploy the financial crimes agent, with a broader rollout planned for the second half of 2026, according to a March 4 release. The agent is expected to reduce AML investigations from hours or days to minutes.
AML is a strong use case because agentic AI streamlines the manual workflows tied to data governance and security, Annie DeStefano, founder and chief executive of Ann DeStefano Advisory, which provides consulting services to financial institutions, told FinAi News.
“It’s certainly an area where the cost of not being smarter or faster around those things is really, really high,” she said. “How this gets deployed in a very controlled manner will be sort of the name of the game across financial institutions.”
Measuring success
According to the release, the new AI agent will be evaluated by metrics including:
- Cost per case;
- Case review times; and
- Reduction of low-value manual work.
Improving the narrative quality of suspicious activity reports is another important metric, Valente said, noting that this “benefits both the institution’s compliance posture and the quality of information reaching regulators.”
The collaboration will combine FIS’ data infrastructure, regulatory expertise and transaction systems with Anthropic’s Claude AI models and embedded engineering teams.
FIS chose Claude because “the use case demands a model that can reason through complex, high-stakes investigations accurately, explain its work and operate safely inside regulated workflows,” Valente said.
Rise of AI agents
The AML agent comes as a flurry of agentic solutions hit the market, with recent adopters of new tools including Citi, ING, Bank of Valletta and Michigan State University Federal Credit Union.
The emergence of agentic AI shows that the technology “has matured to the point where regulated institutions are comfortable enough adopting and bringing it into their infrastructure,” DeStefano said.
Meanwhile, Anthropic and FIS plan to continue co-developing agents to address other pain points in banking, the release states.
Register here for the FinAi Lending Summit, set for Oct. 7-8 in Las Vegas.
