OpenAI launches ChatGPT for Financial Services with Morgan Stanley and Evercore

23 Sep 2026

OpenAI has unveiled ChatGPT for Financial Services, a product built to handle some of the most time-consuming work on Wall Street: researching companies, crunching financial data, and assembling the polished slide decks that investment bankers depend on to win deals.

The product is a tailored version of OpenAI’s existing enterprise tool, ChatGPT Work. It runs on GPT-6 Astra, the company’s latest model, and was developed with design partners Morgan Stanley and Evercore. Nick Turley, OpenAI’s vice president of product, announced it during a briefing and demonstrated it live, showing the platform analysing a potential merger and acquisition target, pulling in financial figures from industry-standard data sources, and producing a formatted PowerPoint deck aligned to a bank’s own style guide.

What ChatGPT for Financial Services actually does

The product’s core selling point is its native access to financial data. Unlike the standard ChatGPT Work tool, this version connects directly to LSEG, Daloopa, Crunchbase, and PitchBook, databases that provide financial statements, earnings transcripts, and company records. It also links automatically to a user’s existing data subscriptions, so analysts do not have to copy figures manually from one system to another.

Other features built specifically for finance include citations that let users trace any data point back to its original filing, tools to audit charts against the underlying numbers, and administrative controls designed for handling sensitive deal materials. Turley drew a comparison to the effect of spreadsheet software on the industry. ‘In the same way that Microsoft Excel transformed the industry and allowed them to produce better analysis faster, you will see technology like this do the same,’ he said.

He also described the ambition behind the product’s research capability: ‘We’re effectively teaching ChatGPT to research like an analyst and back up its conclusions like an analyst as well.’ The initial focus is on investment banking and equity research, though Turley said OpenAI plans to release tailored solutions for ‘a number of sectors’ beyond financial services.

ChatGPT for Financial Services enters a competitive market

OpenAI is not the first company to pitch a specialised AI tool at Wall Street. According to Fortune, Anthropic announced its own equivalent product, Claude for Financial Services, in July 2025. The two launches mark a broader race among AI companies to establish a foothold in the enterprise market, where contracts tend to be large and sticky. OpenAI has spent much of the past year competing with Anthropic and Google for business customers.

The financial health of its design partners adds context to the timing. Evercore, one of the two banks that helped shape ChatGPT for Financial Services, has been expanding its footprint: CB Insights records that Evercore acquired Robey Warshaw on 30 July 2025, a move that extended the firm’s reach into European advisory work. That kind of growth tends to increase demand for the precise analytical and presentation work that OpenAI’s new product targets.

Sarah Friar, OpenAI’s finance chief, told investors in August that the company’s enterprise business had already overtaken its consumer business in revenue terms. ChatGPT for Financial Services represents OpenAI’s clearest attempt yet to deepen that lead, and comes as the company is widely expected to pursue an initial public offering.

The bigger question for junior bankers

The launch lands squarely in an ongoing debate about what AI means for entry-level roles in finance. Wall Street has long relied on a pipeline of recent graduates, known as analysts and associates, to do the research and produce the pitchbooks that senior bankers present to clients. If a tool can complete those multistep tasks in minutes, the industry faces hard questions about how many junior staff it needs, and how it trains them.

When asked directly whether ChatGPT for Financial Services would reduce hiring of junior bankers, Turley framed it as a productivity gain rather than a headcount cut. Analysts, he noted, routinely work 100-hour weeks, and the product is intended to help them work faster, not replace them outright.

Not everyone on Wall Street is reassured. Chris Churchman, the Goldman Sachs partner in charge of one of the bank’s flagship AI projects, warned last month that automating the tasks that train junior bankers risks causing ‘cognitive atrophy’ in the next generation of financiers. ‘Reasoning is still important,’ Churchman said. ‘You still need to reason about [problems] and structure it into an argument, and now we’re delegating reasoning.’

Turley declined to name any banks that have formally signed on for the product, though he said demand had been strong. OpenAI has confirmed it was built in partnership with Morgan Stanley and Evercore, both of which were involved from the design stage.

Brendan Carey

Brendan Carey spent sixteen years in education before he started writing about it. He trained as a secondary school computing teacher, ran an ICT department for a decade, and spent his last years as a digital learning lead at a multi-academy trust trying to get procurement budgets to match curriculum ambitions. He writes about edtech, digital skills, curriculum policy, and the reality of teaching technology in schools where the Wi-Fi drops out during inspections. He knows what works in a classroom and what works in a sales demo, and they are rarely the same thing. Brendan lives in Leeds. He has opinions about the computing curriculum that he considers constructive and Ofsted would consider robust.