Timed to an event Anthropic is hosting in New York, the startup ​launched 10 financially focused agents, or AI programs that carry out tasks with ​limited human intervention. These include agents that can build a pitchbook, ⁠audit statements or draft credit memos, Anthropic said, News.Az reports, citing Reuters.

The company also announced more data sources ​that its so-called Claude AI can access to perform financial work.

Not yet a year into ​unveiling ambitions to tailor AI for finance, Anthropic has rapidly expanded its business, with adoption by Goldman Sachs (GS.N), Visa (V.N), Citi (C.N), AIG (AIG.N), opens new tab and others. Banks have scrambled to access its Claude Mythos model to shore up ​their cybersecurity. At the Tuesday event, Anthropic CEO Dario Amodei was due to appear ​on stage with Jamie Dimon, the chief executive of JPMorgan Chase (JPM.N), opens new tab.

Anthropic's drive to automate enterprise work has ‌meanwhile ⁠hammered an array of financial, legal and software stocks, due to anticipation that the AI provider would disrupt or supplant their businesses. The San Francisco-based AI lab has said it wants to improve outcomes for customers, rather than replace them.

In an interview, Nicholas Lin, who ​leads Anthropic's financial services ​product work, said an ⁠increasingly capable Claude would develop "vertical-specific intelligence," for instance in finance, even as the startup's AI is widely applicable across industries.

These AI model ​advances, coupled with hands-on customer support and key office software integrations, ​were underpinning ⁠a rapid uptick in Anthropic's financial services business, said Lin.

"I've honestly seen a dramatic change, especially in the past six months," Lin said.

As part of the announcement, Anthropic said the ⁠10 new ​AI agents could plug in to its Claude ​Code and Cowork products out of the box, and could be customized to a firm's policies and style.