FIS and Anthropic build financial crimes AI agent for banks

TL;DR:

  • Banking technology vendor FIS and Anthropic are co-developing a Financial Crimes AI Agent that compresses anti-money-laundering investigations from hours to minutes by assembling evidence across a bank’s core systems and surfacing the highest-risk cases.
  • BMO and Amalgamated Bank are among the first deployers, with broader availability planned for H2 2026. FIS sits inside roughly 12% of the world’s banks — a major distribution channel for Anthropic into financial services.
  • Resultsense view: this is the second material Anthropic-finance announcement in a week, after the company’s launch of ten ready-to-run finance agent templates. The architecture — vendor-of-record platform plus model provider — is the format UK banks should expect compliance-grade AI to arrive in.

The Financial Crimes AI Agent will automatically assemble evidence across a bank’s core systems, evaluate activity against known typologies, and surface the highest-risk cases for investigator review. The roadmap also includes credit decisioning, deposit retention, onboarding and fraud agents.

How the partnership is structured

Under the arrangement, FIS provides the foundation: data platform, governance layer, deployment infrastructure and client relationships. Anthropic’s Claude models supply the reasoning. Anthropic’s Applied AI team and forward-deployed engineers are embedded with FIS to co-design the agent and transfer knowledge so FIS can scale additional agents independently over time.

The split matters: FIS is the regulated vendor of record for institutions that need a single accountable counterparty under regulatory regimes like the FCA’s SYSC and the EBA’s outsourcing guidelines. The model provider sits one layer back, behind the platform’s controls and audit trail.

What Stephanie Ferris said

FIS chief executive Stephanie Ferris framed the launch in pointed terms: “Every bank in the world wants AI that acts, not just assists. The future is about a trusted provider who manages the data, who governs the agents, and who stands between your customers and the AI making decisions about their money.” That language — provider sitting “between your customers and the AI” — is exactly the framing UK regulators have been signalling they want to see in agentic finance deployments.

The wider Anthropic-finance push

The FIS deal sits inside Anthropic’s broader same-week push into financial services: ten new agent templates (pitch builder, model builder, KYC screener, valuation reviewer, month-end closer, statement auditor and more), Claude add-ins for Microsoft Excel, PowerPoint, Word and Outlook, and new connectors with Dun & Bradstreet, FactSet, S&P Capital IQ, MSCI, PitchBook, Morningstar and others. Anthropic claims the underlying Claude Opus 4.7 leads Vals AI’s Finance Agent benchmark at 64.37%.

UK relevance

UK retail and wholesale banks running on FIS core systems — among them parts of NatWest’s processing stack and a number of building societies — will see the Financial Crimes AI Agent appear inside their existing vendor relationship rather than as a new procurement. That changes the usual pace of AI adoption in UK banking, where compliance and supplier-risk reviews tend to add 12 months to evaluation cycles. The City and the FCA have been scrutinising agentic AI deployment in financial crime since late 2025, with explicit interest in audit trails for autonomous decision-making.

Looking forward

Watch for whether UK banks already on FIS publish how they plan to govern the agent’s autonomous actions, and how the FCA’s evolving expectations on outsourced AI will treat the FIS-Anthropic arrangement. The architecture — platform vendor plus frontier model provider — is likely to be replicated; rivals such as Temenos, Finastra and Mambu are under direct pressure to announce similar deals.

Share this article