Our Take
The agent catalogue approach sidesteps the hard problem of enterprise data integration by building narrow workflows first.
Why it matters
Mid-market banks get a tested playbook from JPMorgan and Goldman adoption, while FIS partnership brings agents to existing core banking infrastructure.
Do this week
Finance AI teams: Audit your current pitchbook and credit memo workflows before Claude agents launch to identify integration points.
Anthropic ships 10 finance agents, Microsoft integration
Anthropic launched 10 Claude agents for financial workflows including pitchbook writing, earnings analyses, credit memos, loan underwriting, and know-your-customer tasks. The agents work through a Microsoft collaboration that carries context across Excel, PowerPoint, Word, and Outlook as a single agent.
Core banking provider FIS announced it will build AI agents alongside Anthropic, starting with an anti-money laundering case investigation agent. "Every bank in the world wants AI that acts, not just assists," said FIS CEO Stephanie Ferris (company statement).
The launch event featured JPMorgan CEO Jamie Dimon, CIO Lori Beer, Goldman Sachs CIO Marco Argenti, and AIG CEO Peter Zaffino discussing their AI adoption. JPMorgan is "rewiring the bank with AI for speed, security, scalability, efficiency and resiliency," Beer said.
New data partnerships connect Claude to Verisk, Third Bridge, Fiscal AI, D&B, Experian, GLG, Guidepoint and IBISWorld. Moody's embedded its platform as a Claude app for credit ratings and risk analysis.
Big bank adoption creates mid-market roadmap
Major bank technology choices drive industry adoption patterns. "The largest institutions have the ability to explore this space at a scale and with a degree of experimentation that a mid-market bank doesn't necessarily have the funding to be on the leading edge of," said Nate Porter, director at West Monroe.
The regulated nature of banking creates follow-the-leader dynamics. "If there's some leadership from the largest institutions, there's a huge volume of banks that are ready to jump in there," Porter noted.
Goldman's Argenti positioned AI as changing decision-making workflows: "The next wave of AI will help us make better decisions - better risk decisions, better investment decisions, faster and smarter."
Start with document workflows, plan data control
The agent selection targets straightforward use cases first. Credit memos and pitchbooks represent "an exercise of taking a set of information and developing a narrative off of it that is easy to understand and push forward through an approval process," Porter explained.
Data ownership remains the critical challenge. Banks building their own data warehouses and lakes position themselves to "control their own destiny around how they access and use their data" rather than accept prepackaged solutions, Porter said.
The FIS partnership matters because it brings agents to existing core banking infrastructure without requiring banks to rebuild data pipelines. For institutions already on FIS systems, the AML agent provides a tested entry point for agent adoption.