Alerts
-
3 minutes, 13 seconds
OpenAI 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 and 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.
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 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.
OpenAI's enterprise business now generates more revenue than its consumer business, CFO Sarah Friar told investors in August. The company is widely expected to pursue an IPO, and finance is a market where Anthropic's Claude for Financial Services arrived first.
Asked by CNBC whether ChatGPT for Financial Services would reduce junior banker hiring, Nick Turley described it as a productivity tool, comparing it to Excel: faster, better analysis from the same people. The comparison has limits. Excel executed the analyst's reasoning faster; it did not choose the peer set, decide which chart mattered, or explain the sell-off.
That concern is not only external. On an August 24 episode of Goldman Sachs' Exchanges podcast, Chris Churchman warned of "cognitive atrophy" if bankers outsource their reasoning to models, comparing the risk to how GPS eroded people's sense of direction.
Stanford's Digital Economy Lab, in research updated in August, found employment among workers aged 22 to 25 in the occupations most exposed to generative AI sits 19% below where it would be had it tracked their less-exposed peers. Experienced workers show no comparable gap. The researchers attribute the divergence to firms hiring fewer young workers, not layoffs, and find the decline concentrated where AI is automating rather than complementing human work. The Federal Reserve Bank of Dallas reached a consistent conclusion in a January analysis.
For banks, the analyst role's content is up for redesign: less deck assembly, more verifying output and pressure-testing assumptions. Whether banks build that version of the job, or simply hire fewer people, is the decision that matters. As Seb Kirk of GaiaLens put it, nobody learned judgment from reformatting a deck at 2 a.m., but handling the numbers yourself is how you noticed when one was wrong.
8.4K articles
1.6K articles
34 articles
28 articles
Comment