OpenAI's ChatGPT for Financial Services: GPT-6 Astra doing junior bankers' research, models, and pitchbooks

#What happened

On September 10, OpenAI launched ChatGPT for Financial Services, a finance-tailored version of ChatGPT Work powered by GPT-6 Astra and developed with Morgan Stanley and Evercore as design partners. It starts with investment banking and equity research. OpenAI bundles and hosts premium datasets itself: Daloopa, PitchBook, LSEG News, and Crunchbase, covering earnings transcripts, financial statements, company fundamentals, and private-company data, with citations tracing figures back to source filings. In a live demo, VP of product Nick Turley had the tool analyze a potential M&A target, pull financial figures from industry-standard sources, select comparable companies, and generate a formatted PowerPoint deck matching a bank's own template. Turley's pitch: "We're effectively teaching ChatGPT to research like an analyst and back up its conclusions like an analyst as well." Admin controls let firms publish Excel, Word, and PowerPoint templates and manage sensitive deal materials. Turley frames it as taking the grind off analysts working 100-hour weeks, not replacing them.

#Why it matters

It is OpenAI's most direct push into a single enterprise vertical, and it lands against Anthropic's Claude for Financial Services (pre-built MCP connectors to FactSet, S&P Capital IQ, Morningstar, PitchBook, Daloopa, Moody's), squeezing startups like Rogo and Hebbia that sell AI research tools to banks. The open question is training: Goldman Sachs partner Chris Churchman warns that automating the grunt work junior bankers learn the job on could leave them worse analysts down the road.