OpenAI on Thursday introduced ChatGPT for Financial Services , a version of its enterprise product built to research companies, analyze financial data and produce the client presentations that investment banks run on. The tasks it performs out of the box — comparable-company analysis, LBO modeling, buyer screening, earnings analysis, pitchbook preparation — are the tasks Wall Street has assigned to its youngest employees for decades.

The product is a tailored build of ChatGPT Work developed with Morgan Stanley and Evercore as design partners, according to Nick Turley, OpenAI’s vice president of product. It runs on GPT-6 Astra, the company’s newest model.

What Is Different About This Version

The distinction from the base enterprise product is data. ChatGPT for Financial Services ships with native access to financial statements, earnings transcripts, company fundamentals and private-company data from LSEG, Daloopa, PitchBook and Crunchbase , plus automated access to a firm’s existing data subscriptions.

Every figure carries a citation that traces back to the source filing, charts can be audited against the underlying numbers, and administrators get controls for sensitive deal materials.

In a live demonstration, Turley had the system evaluate a potential M&A target, pull the relevant figures and produce a formatted PowerPoint deck in the bank’s own template. The hard part, he told reporters, was not making slides that look good but making slides that make sense: the model had to select the right peer set, load prices into a spreadsheet, verify the chart against the data and explain a sell-off and rebound.

Turley said demand for the product was strong and that OpenAI plans similar sector-specific versions beyond finance. He declined to name any banks that have signed on.

The launch fits a pattern. OpenAI’s chief financial officer, Sarah Friar, told investors in August that the company’s enterprise business now generates more revenue than its consumer business. The company is widely expected to pursue an initial public offering, and finance is a market where its rivals arrived first: Anthropic released Claude for Financial Services last year.

Banks are among the highest-value customers in software. They pay for data, they pay for reliability, and they employ tens of thousands of people whose daily work maps closely to what large language models do well. That combination makes investment banking a natural first vertical — and a revealing one.

The Question OpenAI Did Not Answer Directly

Asked by CNBC whether the product would reduce the need to hire junior bankers, Turley described it as a productivity tool rather than a replacement. Analysts work 100-hour weeks, he said, and the effect would resemble what Microsoft Excel did for the industry: faster, better analysis from the same people.

The comparison is worth examining, because the two technologies differ in one important respect. Excel executed the analyst’s reasoning faster. It did not choose the peer set, decide which chart mattered or explain the sell-off. A tool that performs those steps is doing the part of the job that trained the analyst in the first place.

That concern is not coming only from outside the industry. On an August 24 episode of Goldman Sachs’ Exchanges podcast , Chris Churchman — a partner who leads the bank’s Marquee platform and co-chairs its Global Banking and Markets AI working group — warned of what he called “cognitive atrophy” if bankers outsource their reasoning to models.

Much of the knowledge in finance is tacit and learned by doing, he said, and he compared the risk to the way GPS eroded people’s sense of direction. Goldman, he acknowledged, has not settled how much of that early-career work it will automate.

What The Labor Data Shows

The apprenticeship question has a measurable early signal. Research from Stanford’s Digital Economy Lab , updated in August, finds that employment among workers aged 22 to 25 in the occupations most exposed to generative AI now sits 19% below where it would be had it tracked their less-exposed peers.

Experienced workers in the same occupations show no comparable gap. The researchers attribute the divergence to firms hiring fewer young workers, not to layoffs of those already employed, and they find the decline concentrated where AI use is automating rather than complementary.

The Federal Reserve Bank of Dallas reached a consistent conclusion in a January analysis : employment losses tied to AI exposure appear only among younger workers, and they are driven mainly by fewer people entering those occupations from outside the labor force.

Investment banking is a specific case within that broader pattern, and the pattern is what the OpenAI launch makes concrete. If a model can produce a defensible pitchbook in minutes, the economic logic of a two-year analyst program — in which a bank pays a recent graduate to learn by producing that same pitchbook slowly, under supervision — changes.

Where Wall Street Goes From Here

None of this means the analyst role disappears. It means the role’s content is up for redesign. The plausible future is one in which junior bankers spend less time assembling decks and more time verifying model output, pressure-testing assumptions and sitting in on the client conversations that used to be reserved for their seniors. Whether banks actually build that version of the job, or simply hire fewer people, is the decision that matters.

Seb Kirk, chief executive of the AI platform GaiaLens, put the stakes plainly in an interview with Moneywise : nobody learned judgment from reformatting a deck at 2 a.m., but handling the numbers yourself is how you noticed when one of them was wrong.

OpenAI has built a tool that handles the numbers. What the industry does with the people who used to is the story that will play out over the next hiring cycle.