ChatGPT goes to Wall Street

OpenAI launched ChatGPT for Financial Services on Sept. 10, a version of ChatGPT Work designed with Morgan Stanley and Evercore and running on GPT-6 Astra.

Built to think like an analyst

It aims at LBO modeling, buyer screening, earnings analysis and pitchbooks, with citations tracing every figure back to the source filing.

“We're effectively teaching ChatGPT to research like an analyst and back up its conclusions like an analyst as well.”

Anthropic shipped this first

Claude for Financial Services arrived May 5 with connectors to FactSet, S&P Capital IQ, Moody's and more, plus 10 agent templates for pitchbooks, KYC and month-end close.

Two bets on the data

Anthropic connects out to each provider's own system through MCP. OpenAI licenses and hosts Daloopa, PitchBook and LSEG News itself, keeping tighter control over retrieval and citations.

A tiered integration model

Bundled datasets need no separate contract or connector. Others, including MSCI, Dow Jones Factiva and Moody's, get a lighter sign-in that just recognizes a bank's existing entitlements.

Startups feel the squeeze

Rogo and Hebbia sell the same thing and run on these very models. Rogo passed $50 million in ARR, up from $15 million at the end of 2025, and is tracking toward $100 million.

Banks are the bigger threat

Rogo and Hebbia argue their agents behave like your firm, not a generic chatbot. But their own customers are building in-house agents, even if most still lack scoped use cases.

“increasingly like your firm rather than like a generic Claude”