Skip to content
NewsProduct launch

OpenAI launches ChatGPT for Financial Services with Morgan Stanley and Evercore as design partners

· by Pondero Newsdesk

The short version

OpenAI's ChatGPT for Financial Services connects GPT-6 Astra to live data from Daloopa, PitchBook, LSEG News, Bloomberg, and FactSet via roughly 50 MCP connectors. Access requires a ChatGPT Enterprise account with direct OpenAI clearance.

OpenAI launches ChatGPT for Financial Services with Morgan Stanley and Evercore as design partners

Six live financial data feeds, a direct line to the GPT-6 Astra model, and a tight access gate: OpenAI on September 10 launched ChatGPT for Financial Services, a product that targets the company research, financial-model building, and pitchbook creation work that junior investment bankers and equity analysts currently own.

What

Morgan Stanley and Evercore shaped the product as design partners, per Fortune. Nick Turley, VP and head of ChatGPT, told Fortune that the collaboration was designed to close the gap between demo-quality output and production-usable work: "There's a difference between what looks good in a demo and what is actually a usable output."

Data connections cover Daloopa (earnings transcripts and financial statements), PitchBook (private-company fundamentals), Crunchbase, LSEG News, Bloomberg, and FactSet. Roughly 50 total connectors are available through MCP, the open-source protocol that lets AI models reach third-party software and data sources, per Fortune. On the output side, the tool generates PowerPoint presentations, Excel spreadsheets, and web-based dashboards. It can also run acquisition analyses with peer comparisons and model how revenue growth or shifting market conditions move valuations.

Access is restricted. Only financial institutions holding ChatGPT Enterprise accounts that receive direct clearance from OpenAI can use the product, per Fortune. A parallel channel for the same underlying model runs through Microsoft Copilot for Finance on Azure OpenAI Service, per CNBC.

Why it matters

The product targets a specific cost center in investment banking: the analyst hours spent on first-draft materials. Pitchbook slides, comps tables, and earnings summaries are not judgment calls; they are structured assembly work. ChatGPT for Financial Services does not replace the senior banker reviewing those outputs, but it can compress the time required to produce them. Fewer hours per deliverable, or more deliverables per analyst headcount, is the operational shift that banks will weigh against vendor risk and compliance review.

The competitive context is direct. Anthropic launched Claude for Financial Services in 2025, making it the incumbent in the credentialed-enterprise-AI-for-finance tier. This gives procurement teams at banks a second lab option with its own named institutional backers. The vetting overhead stays high on both sides: neither OpenAI nor Anthropic offers open enrollment for their financial-services products. But having two credentialed alternatives changes how banks negotiate.

What to watch next

Two things will show whether the product gains traction beyond its design partners. Public commitments from Goldman Sachs or JPMorgan would signal that the enterprise clearance process is moving at scale. The other question is positioning: Bloomberg Terminal now appears as one of six data connectors in a competing product. How Bloomberg responds at the workstation level, where the Terminal's own AI layer competes directly, will shape how financial professionals choose between a vertically integrated terminal and a chat-layer product with broker-neutral data access.

Sources