The day following the $1.5 billion joint venture, Anthropic delivered the products that the JV will market.

      The New York event on Tuesday unveiled Claude Opus 4.7, a library containing around ten pre-built finance agents, alongside an AML investigator developed by FIS that will be deployed at BMO and Amalgamated Bank, as well as a Moody's native app covering 600 million companies. This followed Monday's announcement of a $1.5 billion joint venture on Wall Street, allowing the product side to further develop.

      On Monday, Anthropic revealed a $1.5 billion enterprise services joint venture with Blackstone, Hellman & Friedman, and Goldman Sachs, marking one of the most prominent AI services investments led by Wall Street to date. By Tuesday morning, the product side had caught up, introducing a collection of pre-built AI agents tailored for the most labor-intensive workflows in financial services, as well as a new model release, a partnership with Moody's, and full integration with Microsoft 365.

      These two announcements signify Anthropic's shift from being a frontier-model lab to a financial-services platform, with both components being essential: the joint venture enables distribution, while the product launch provides the tools to be distributed.

      The main highlight of Tuesday's announcement, which took place at Anthropic’s exclusive “Briefing: Financial Services” event in New York, was Claude Opus 4.7, the company's most advanced model for finance-related tasks. It comes with a library of approximately ten pre-built agents that address the workflows consuming the most time for analysts in banking and asset management, including pitchbooks, earnings analysis, credit memos, KYC, month-end close, statement audits, and insurance claims, among others. Each agent is provided as a reference architecture, complete with necessary skills, connectors, and sub-agents for seamless workflow execution.

      Importantly, these agents are not generic; they are pre-configured to integrate with the data sources commonly used by finance teams. Moody's is incorporating its entire platform into Claude as a native application, enabling users to extract credit ratings, risk data, and ownership-structure analysis for over 600 million companies directly within the Claude interface.

      Companies such as Verisk, Third Bridge, Fiscal AI, Dun & Bradstreet, Experian, GLG, Guidepoint, and IBISWorld have joined an existing roster of data partners that already featured LSEG, S&P Capital IQ, Morningstar, and PitchBook, essentially covering much of the relevant equity-research and credit-analysis data.

      The partnership with FIS is noteworthy. FIS, a publicly traded banking technology company, announced on Monday that it has developed a Financial Crimes AI Agent using Claude, aimed at reducing anti-money-laundering investigations from hours or days to mere minutes. BMO and Amalgamated Bank are the first announced users, with wider availability expected in the latter half of 2026.

      What sets this apart from a chat product is significant. Anthropic has previously released financial-services tools, with earlier iterations of Claude for Financial Services focusing on plugins that enabled analysts to query financial data within the Claude chat interface. Earlier this year, the collaboration with Xero allowed Claude to present small-business accounting data.

      However, Tuesday's announcement represents a structural shift, transitioning Claude from a tool that analysts use to pose questions into a system capable of autonomously executing predefined workflows, complete with audit trails, regulatory compliance, and embedded governance.

      Claude Managed Agents, the underlying platform launched by Anthropic in April, simplified the technical complexities of deploying autonomous agents by transferring the orchestration responsibilities to Anthropic’s model instead of requiring customer engineering teams. This financial-services suite is the first large-scale, vertically-specific deployment of this platform.

      The orchestration is handled by the agents, while the banks supply the data and governance, effectively positioning Anthropic as the operating system.

      Details of the FIS collaboration reveal its significance. The Financial Crimes AI Agent directly addresses a regulatory issue—AML—where the labor costs of compliance have escalated faster than almost every other expense line in bank operating budgets. FIS estimates the global financial-crimes challenge carries annual costs of about $40 billion in fines, investigations, and false-positive reviews.

      If it proves effective at scale, reducing AML investigation time from hours to minutes would greatly benefit any chief compliance officer at a large bank. BMO has been named as a launch partner for this initiative; earlier, we explored BMO's broader AI and quantum strategy, highlighting its proactive stance as a visible early adopter of a Claude-based AML agent.

      The structure of the partnership illustrates the model's collaborative nature. FIS contributed the necessary financial-data infrastructure, regulatory connectors, and bank-system integration, while Anthropic provided Claude's reasoning capabilities, agent orchestration, and embedded engineering support.

      Evidently, Anthropic is targeting what it perceives as a highly lucrative opportunity within enterprise AI. Financial services represent the largest sector in the global economy in professional services, with revenues from consulting, audits, and advisory services in leading firms reaching tens of billions in each