Anthropic Launches AI Agents for Finance Amid Investments and Pentagon Exclusion

Anthropic has launched 10 new AI agents aimed at streamlining workflows in the financial services industry. The tools, announced on Tuesday, include features like a model builder for creating financial models from filings and analyst notes, and a pitch builder for drafting pitchbooks ahead of client meetings.

Financial services represent Anthropic's second-largest industry by enterprise revenue, with 40% of its top 50 customers in finance. Startups like Rogo and Hebbia are also developing similar tools, with Rogo serving over 250 clients and emphasizing model-agnostic workflows backed by finance expertise.

EY's financial services technology consulting leader noted that differentiation will shift to domain-specific data and integration with existing systems.

Google parent Alphabet reported an 81% profit jump to $62.6 billion in the first quarter, with nearly half—about $28.7 billion—stemming from unrealized gains on its equity stake in Anthropic. Amazon disclosed $16.8 billion in pre-tax gains from its Anthropic investment, accounting for more than half of its quarterly pre-tax income.

These markups followed Anthropic's recent funding round, which valued Amazon's $8 billion investment at over $70 billion. The gains reflect accounting rules requiring updates to private equity valuations based on new funding rounds. A tax and accounting consultant highlighted that such markups can introduce volatility to earnings, as they depend on investor agreements rather than open market prices.

Both companies have committed additional billions to Anthropic, separate from commercial relationships. The partnerships aim to establish the U.S. military as an AI-first force and maintain decision superiority in warfare. The agreements cover Impact Level 6 and 7 environments for lawful operational use.

Anthropic was notably absent from the deals, following its designation as a supply-chain risk in March due to refusal to grant unrestricted access to its Claude models for autonomous weapons and surveillance. The company filed lawsuits in San Francisco and Washington, D.C., to overturn the order, with the case ongoing.

Recent meetings between administration officials and Anthropic's CEO suggest potential for resolution. Advancements in AI are raising concerns about workforce impacts in finance, with some bank CEOs mentioning slowed hiring and redeployment plans for roles affected by automation.

No mass layoffs have been announced, but tools are expected to automate tedious tasks across junior and engineering roles. In defense, the exclusions underscore tensions over AI ethics and national security applications.

Anthropic's entry intensifies competition in AI tools for Wall Street, where banks and startups are automating grunt work like market research and earnings reviews. The 10 agents also include a meeting preparer, valuation reviewer, and KYC screener, among others.

As generative AI evolves, products that integrate seamlessly with governance and risk frameworks are positioned for success, according to industry consultants. Blackstone and Goldman Sachs are among investors in a new firm to integrate Anthropic's Claude model into their systems.

This follows Anthropic's presentation highlighting finance as a key revenue driver. The developments come amid broader AI adoption, with tech giants' profits heavily influenced by their stakes in the startup.