Anthropic launched a set of AI agents targeting financial services professionals on Tuesday, releasing tools designed to handle tasks ranging from building pitch materials to running compliance checks on client files.
The San Francisco-based startup released 10 agent templates built for work across research, operations, and finance functions.
On the research side, the new agents include a pitch builder that generates comps models and draft pitchbooks, a meeting prep tool, an earnings reviewer that reads transcripts and flags relevant model updates, and a financial model builder.
Operations-focused templates include a general ledger reconciler, a month-end closer, a financial statement auditor, and a KYC screener that assembles entity files and packages escalations for compliance teams.
The agents are designed to deploy in two ways: as plugins inside Claude Cowork or Claude Code, where they run alongside an analyst's existing desktop software, or as Claude Managed Agents on the Claude Platform, which can handle longer-running tasks autonomously – including multi-hour deal processes – with full audit logs accessible to compliance and engineering teams.
In both cases, users retain approval authority over Claude's work before anything reaches a client or gets filed, according to the announcement on Anthropic's website.
The company also revealed that Claude now works across Microsoft Excel, PowerPoint, and Word through add-ins for Microsoft 365, with Outlook support coming later. Once installed, context carries between applications automatically – a financial model started in Excel, for instance, does not need to be re-explained when moving to a PowerPoint deck.
Nicholas Lin, head of product for financial services at Anthropic, said the company has invested in reinforcement learning specific to finance topics – a technique used to train agents to optimize their results. That training does not include any client data, he said.
"It's important for us to build the model capabilities for the next six months, not today," Lin said in an interview with Barron's, adding that client feedback shapes that process. "This is how we really shape the exponential, as we call it. Once Claude is able to do something – even if not well – we're able to get feedback on how good it is at doing that task, and then we bring that back to the model."
The agent release came one day after Anthropic revealed a joint venture with Blackstone, private-equity firm Hellman & Friedman, and Goldman Sachs. Each of the three anchor investors are expected to kick in roughly $300 million, and Goldman Sachs is contributing around $150 million. General Atlantic, Leonard Green, Apollo Global Management, GIC, and Sequoia Capital are also participating. The total commitment is expected to reach approximately $1.5 billion.
The as-yet-unnamed entity is expected to act as a consulting arm, helping businesses – including portfolio companies of the participating private-equity firms – integrate AI across their operations.
OpenAI is pursuing a similar arrangement. Bloomberg News reported Monday that the ChatGPT maker is finalizing a separate joint venture with private-equity firms to expand adoption of its own tools. Both companies are targeting corporate clients generally, and companies backed by private equity in particular, given that segment's emphasis on efficiency and cost reduction.
Anthropic also expanded its data partner ecosystem Tuesday, adding connectors to Dun & Bradstreet, Fiscal AI, Financial Modeling Prep, Guidepoint, IBISWorld, SS&C IntraLinks, Third Bridge, and Verisk. Those connectors give the agents real-time, governed access to market data, expert interview transcripts, industry financials, insurance data, and deal room documents.
Moody's launched a separate MCP app that surfaces proprietary credit ratings and data on more than 600 million public and private companies.
The moves extend a push Anthropic began last year, when it launched a broader suite of tools aimed at banks, financial-technology firms, and others. That effort, along with strong performance from its Claude Code developer product, has contributed to a significant revenue surge at the company, which is among a handful of names pursuing trillion dollar-plus IPOs this year.
Anthropic was last valued at $380 billion; OpenAI closed its most recent funding round in March at a valuation of $852 billion.
Choice anxiety, prestige bias, and the temptation to make selections based on outsourced confidence are just some of the parallels between investing and the world of wine tasting.
Regulators found Bank of America's monitoring software had a known flaw Merrill left uncorrected for years.
While AI has become a go-to research tool for affluent investors, new HSBC research suggests human advisors remain the deciding voice when investment decisions are made.
A 5-4 ruling preserves the Federal Reserve's independence for now, but the legal fight over presidential removal power is far from settled.
For years, large firms have been facing penalties and questions from regulators over interest rates for clients’ cash accounts.
Dan Biagini of American Equity says the steady decline of pensions, longer lifespans and a reset in interest rates are rewriting how advisors build retirement income
Direct indexing is on pace to outgrow ETFs and mutual funds. Northern Trust's Ken Lassner explains why the advisors who get it wish they had started sooner.
