2025-10-27 · Anthropic

Advancing Claude for Financial Services

agentsmodelsenterprisecapital

read at source ↗ www.anthropic.com

Advancing Claude for Financial Services

Source: Anthropic Date: 2025-10-27 URL: https://www.anthropic.com/news/advancing-claude-for-financial-services

Summary

Anthropic expanded its financial services vertical with Claude for Excel (beta sidebar integration for Max/Enterprise/Teams), real-time market data connectors to Aiera, Chronograph, Egnyte, LSEG, Moody’s, and MT Newswires, and six Agent Skills for financial tasks including DCF modeling and coverage report generation. Sonnet 4.5 was cited as topping the Vals AI Finance Agent benchmark at 55.3% accuracy. Endorsements from Citi, RBC Capital Markets, Brex, Block, and Coinbase anchored the announcement.

Implications

  • Enterprise positioning thread. Financial services is Anthropic’s highest-value vertical play — complex documents, high compliance bar, large LTV customers. The Excel sidebar is a direct distribution move into a workflow OpenAI already targets with Copilot.
  • Agent Skills as moat. Domain-specific skills (DCF modeling, coverage reports) are stickier than raw model capability because they encode workflows. Building these into the platform raises switching costs.
  • Benchmark as sales tool. Leading with a vertical benchmark (Vals AI Finance Agent, 55.3%) is a departure from general capability claims — Anthropic is adapting its go-to-market to sector-specific evaluation, which will pressure competitors to match on the same benchmarks.
  • Watch: whether the Excel beta converts to GA; adoption signals from named enterprise customers; how OpenAI’s Copilot for Finance responds.

← all signals