OpenAI built a ChatGPT for bankers, and the part that stands out isn’t the model. It’s the data in the box. ChatGPT for Financial Services, announced September 10, comes with datasets from Daloopa, PitchBook, LSEG News and Crunchbase already loaded and hosted by OpenAI, no contracts to negotiate, no connectors to wire up, and every figure carries a citation back to the table it came from.
Morgan Stanley and Evercore shaped it, and it starts where their people live: investment banking and equity research, where OpenAI says getting at reliable data and turning out polished documents were the biggest headaches. If your firm already pays for S&P Capital IQ, LSEG, MSCI, Factiva or Moody’s, it recognizes you through your ChatGPT login. Admins load the firm’s Excel, Word and PowerPoint templates so the output comes out in house style, compliance can pull the logs into their audit tools, and separate workspaces keep the information barriers up.
It’s sold to “eligible financial institutions,” through sales, and there’s no price on the page. The use cases OpenAI lists are deal work: value analysis, LBO modelling, buyer screening, earnings analysis, pitchbooks. Whether a two-person advisory firm here qualifies isn’t a question the page answers, and it’s one we’d like to put to OpenAI.