FIS and Anthropic Build AI Agent to Fight Financial Crime
FIS has partnered with Anthropic to bring agentic artificial intelligence into banking, with the first product squarely aimed at one of the industry's most stubborn cost centres: financial crime compliance. The two companies are jointly developing a Financial Crimes AI Agent designed to compress anti-money-laundering investigations from days into minutes.
The economics behind the move are hard to ignore. The United Nations estimates that roughly $2 trillion in illicit funds flows through the global financial system every year, while US financial institutions spend an estimated $35 to $40 billion annually on AML operations. Much of that spend goes on manual investigative work that the new agent is built to absorb.
How the agent works
Rather than replacing investigators, the Financial Crimes AI Agent automates the labour-intensive groundwork that precedes a decision. It assembles complete evidence packages across a bank's core systems through secure connections, evaluates activity against known financial crime typologies, and surfaces the highest-risk cases for human review. It can also draft Suspicious Activity Reports, leaving investigators to focus on judgement rather than data gathering.
Governance sits at the centre of the design. Client data remains within FIS-controlled infrastructure, every agent action is traceable and auditable, and human investigators retain final decision-making authority. To accelerate the build, Anthropic's Applied AI team and forward-deployed engineers are embedded directly within FIS teams, transferring knowledge so that FIS can develop and scale further agents independently over time.
A platform play
The financial crime agent is positioned as the opening move in a broader strategy. FIS, which says it powers nearly 12% of the global economy, plans to release a suite of curated agents on the same platform targeting credit decisioning, deposit retention, customer onboarding and fraud prevention. BMO and Amalgamated Bank are among the first institutions lined up to deploy the agent, with general availability planned for the second half of 2026.
"The future is about a trusted provider who manages the data, governs agents, and stands between customers and AI making financial decisions," said Stephanie Ferris, CEO and President of FIS.
"FIS chose Claude because they needed a model that could reason through complex investigations accurately and operate safely inside regulated workflows," said Jonathan Pelosi, Head of Financial Services at Anthropic.
For banks weighing the practical use of generative AI, the partnership is a notable signal. Rather than a customer-facing chatbot, it targets a regulated, high-stakes back-office function where speed and auditability both matter, and where the cost of getting compliance wrong is measured in fines and reputational damage.
