OpenAI Targets Wall Street Banks With a Finance-Specific Version of ChatGPT

The product combines financial data sources with the reasoning capabilities of GPT-6 Astra and is intended to help financial teams conduct research, build models and produce branded client materials. OpenAI Vice President and Head of ChatGPT Nick Turley told reporters that financial services is among the verticals the company is prioritizing most, alongside software engineering and cybersecurity.

OpenAI is entering a market where its chief rival has already staked a claim. Anthropic PBC launched a dedicated financial services product, Claude for Financial Analysis, in May 2025. Turley said OpenAI expects its own offering to be comprehensive enough to become the "one product" that banks and other financial firms with thousands of workers need. "This is the canonical product we are hoping the industry adopts," he said.

Turley said he played a large role in developing the product and spent time with several large financial institutions, including Morgan Stanley and the investment advisory firm Evercore Inc., to work out which features financial teams require and to make the tool reliable. "There's a big difference between what looks good in a demo and what is actually a usable output, [and] you kind of rely on the experts to achieve that," he said.

Access requires an enterprise subscription and a direct application to OpenAI, because the product is currently limited to "eligible institutions." Turley said it resembles ChatGPT Work in many ways, with features derived from that tool but given a "finance-specific bend." The interface looks like standard ChatGPT with extra toggles for connecting financial data. Banks can link their own data or datasets from providers such as Bloomberg and FactSet, and OpenAI offers pre-loaded datasets from Crunchbase, Pitchbook, Daloopa and LSEG News. About 50 Model Context Protocol connectors allow links to other data and third-party software. When such data informs an output, the product provides citations, Turley said.

Users can also set the level of effort ChatGPT applies when generating a response: high, medium or low. Higher effort consumes more tokens and produces slower responses, but should mean better quality. Beyond answering financial questions, the tool can create PowerPoint slides, Excel spreadsheets and web-based dashboards.

In a blog post, OpenAI said users can conduct deep research across multiple datasets and produce "detailed artifacts" from a single prompt. For an acquisition, the company said, they could compare a target with its peers, test how revenue growth affects valuation and turn the analysis into an editable model or pitchbook using their firm's templates.

Turley said a broader goal is to change how financial analysts work. A dashboard generated by ChatGPT could let an analyst simulate how different conditions affect forecasts by entering the condition and generating a revised forecast, rather than editing numbers in a spreadsheet. "We're trying to think of new ways of doing the work, rather than just making the existing ways faster," he said.

The automation has also drawn attention to its effect on entry-level hiring. If a chatbot can complete research and build a pitchbook in minutes, banks may be less likely to assign such tasks to junior staff, in an industry where workers traditionally undergo a rigorous apprenticeship. Goldman Sachs partner Chris Churchman, who heads the bank's flagship AI projects, raised that concern in an interview with CNBC last month, according to SiliconANGLE.