Anthropic announced on Wednesday the launch of 10 ready-to-use financial AI agent templates. Not only do they integrate the full suite of Microsoft 365 office applications, but they also bring in multiple financial data partners such as Moody’s and Dun & Bradstreet to comprehensively strengthen Claude’s hands-on capabilities in the finance sector.
10 Financial AI Agents, covering investment banking, compliance, and finance operations
For this update, Anthropic released 10 out-of-the-box AI agent templates specifically for the financial services industry. Each template integrates three major modules—Skills, data connectors (Connectors), and subagents (Subagents)—enabling financial institutions to deploy and go live within days. Each template can be used not only as a plugin for Claude Cowork or Claude Code, but also to run automated scheduling through Claude Managed Agents.
The 10 agents are divided into two major categories, including:
Research and customer service:
Pitch builder: Automatically creates a list of target customers, executes company comparison and analysis, and drafts an investment presentation (pitchbook)
Meeting preparer: Automatically compiles background summaries of clients and counterparties before meetings
Earnings reviewer: Reads financial reports and regulatory documents, updates financial models, and flags major changes related to investment theses
Model builder: Builds and maintains financial models based on financial reports, data sources, and analyst inputs
Market researcher: Tracks industry and issuer developments, integrating news, financial reports, and broker research reports
Finance and operations management:
Valuation reviewer: Automatically reviews valuation results based on comparable companies, valuation methodology, and company approval standards
General ledger reconciler: Reconciles general ledger accounts and performs net asset value (NAV) calculations
Month-end closer: Executes month-end closing checklists, prepares journal entries, and produces closing reports
Statement auditor: Reviews the consistency, completeness, and audit-readiness status of financial statements
KYC screener: Organizes entity record files, reviews source documents, and compiles escalation cases that need to be reported to compliance departments
Of course, before users submit reports, archive documents, or perform any real actions, they can review, modify, and approve Claude’s output results end to end.
Deeply integrated with Microsoft 365, automatically extending across application contexts
Another major highlight of this Anthropic update is that Claude has moved into Microsoft Excel, PowerPoint, Word, and the soon-to-launch Outlook via Add-ins. Claude’s knowledge and work context can now automatically carry over among four applications.
In other words, analysts can use Outlook to automatically organize the inbox, schedule meetings, and reply to emails; build and analyze financial model work in Excel based on financial reports; and without needing to restate everything, directly create investment presentations in PowerPoint.
Expanding the financial data ecosystem, credit agency Moody’s also becomes a partner
The usefulness of AI agents depends on the quality and breadth of data they can access. Anthropic has added multiple new partners to provide real-time data integrations, including Dun & Bradstreet for global enterprise identity verification, Fiscal AI for real-time stock fundamentals, Financial Modeling Prep for cross-asset real-time quotes, and IBISWorld for cross-enterprise financial data.
In addition, credit agency Moody’s has gone further by launching MCP applications, embedding its proprietary credit rating data for more than 600 million companies directly into the Claude interface for use in compliance analysis and business development.
Adopted by top financial institutions, leading in industry benchmark testing
Anthropic says the best combination model for this update is Claude Opus 4.7, which ranked first in the Vals AI Finance Agent Benchmark (Finance Agent Benchmark) with a score of 64.37%. Currently, multiple firms—including Citadel, BNY (Bank of New York Mellon), and Mizuho Financial Group—covering banks, asset management, insurance, and hedge funds, have all said they are actively adopting it.
Claude’s wave reaches the finance industry; regulatory progress and talent structure face challenges
The significance of Anthropic’s deep foray into financial workflows goes far beyond efficiency gains at the tool level. Finance has long relied on people-stacked, information-intensive work, including due diligence, compliance screening, and financial modeling, and this is being rapidly replaced by agents. This shift will undoubtedly reshape the talent demand structure: senior capabilities to review, judge, and correct AI outputs will become increasingly critical, while entry-level roles that only carry out repetitive analysis will face pressure.
At the same time, as AI agents begin to take part in credit reviews and compliance tasks with legal effect, the gray area of accountability will also force regulators in each country to accelerate the formulation of corresponding rules. Ultimately, the outcome of this race will depend on how quickly financial institutions adopt AI and whether regulatory frameworks can keep pace with technological evolution.
(Anthropic Mythos is too powerful! The White House plans to require government security reviews before new AI models are released)
This article about Anthropic launching 10 financial AI agents, integrating with Microsoft 365 to easily handle finance operations, first appeared on Chain News ABMedia.
