Anthropic has expanded its financial services offerings, launching 10 new AI agents designed to automate and accelerate complex tasks for banks, insurers, and other financial institutions. These agents integrate directly into existing Claude products, such as Claude Code and Claude Cowork, to streamline workflows like credit underwriting, risk assessment, and financial modeling. This strategic move aims to embed Anthropic deeper into Wall Street operations, positioning the company for significant enterprise adoption. The release helps bridge the gap between rapid AI development and its practical application within financial firms, according to Jonathan Pelosi, Anthropic's Head of Financial Services, as reported by The Wall Street Journal. This effort is part of a broader push to solidify Anthropic's market presence and accelerate commercialization ahead of an anticipated initial public offering (IPO) later this year. Key partners already adopting Claude include Goldman Sachs, Visa, Citi, and AIG.
What Tasks Do These New Agents Automate?
The new agents focus on critical financial workflows, integrating with widely used applications like Microsoft 365. They enable professionals to build pitch books and financial models in PowerPoint and Excel, draft credit memos, or conduct Know Your Customer (KYC) screening. These tools are designed to handle functions from due diligence to complex analysis.For instance, Stephanie Ferris, CEO and President of FIS, highlighted that her company is building an AI agent with Anthropic that can compress Anti-Money Laundering (AML) investigations from days to minutes . This illustrates the potential for significant time savings in compliance-heavy operations. Similarly, commercial bankers can use Claude to draft credit memos and underwriting models directly in Excel, while asset managers can generate performance summaries.
How Does Claude Ensure Financial Compliance and Security?
Financial institutions demand high standards for security, compliance, and data integrity. Anthropic emphasizes that Claude's models are "source-attributed," meaning every output can be traced back to its data origin. This feature helps teams verify work before acting on it. The platform also meets stringent compliance requirements, including SOC 2 and FedRAMP, which are crucial for risk committees in the financial sector, according toAnthropic.This focus on verifiable data is further enhanced through integrations with trusted data providers. Anthropic has expanded technical partnerships with firms like Dun & Bradstreet, Moody's, Morningstar, and FactSet. These collaborations allow Claude to reason over foundational market data, research, and analytics within secure environments.








