OpenAI introduced ChatGPT for Financial Services on September 10, 2026, positioning the product as a dedicated workspace for investment banking and equity research teams. According to Reuters, Morgan Stanley and Evercore participated as design partners during development. Their involvement indicates that the service was shaped around real institutional workflows rather than offered as a simple rebranding of a general chatbot. The product combines financial datasets, company-controlled templates and enterprise security features in one environment intended to support research, modeling and client-material production.
The service is powered by OpenAI’s GPT-6 Astra model. OpenAI says the model was designed to improve retrieval across financial data tools, financial reasoning and the accuracy of generated material. Teams can search across several sources, build financial models and prepare items such as pitchbooks with their own firm templates. The practical objective is to reduce the time employees spend moving among terminals, documents and separate information services while keeping the work connected to identifiable sources.
Included datasets come from providers such as Daloopa, LSEG News, PitchBook, Crunchbase and Quartr. They cover areas including earnings-call transcripts, financial statements, company fundamentals and firmographic information. OpenAI says this material is indexed on its infrastructure to improve retrieval and citation. Organizations that already subscribe to other services can also connect eligible data through integrations with FactSet, S&P Global, Preqin and Datasite. The arrangement does not erase the underlying licensing boundaries: each institution still needs the appropriate permissions for the information it uses.
Governance is a central part of the launch because financial firms operate under strict controls for confidential information, record keeping and employee access. ChatGPT for Financial Services builds on ChatGPT Enterprise protections including encryption and role-based access. Compliance teams can export workspace logs into audit processes, according to the company’s description reported by Reuters. These controls are intended to help institutions review how the system is used, although they do not remove the need for internal approval procedures or expert supervision.
The launch also reflects a wider shift toward industry-specific artificial intelligence products. A broad model may be capable of answering financial questions, but regulated institutions require more than general capability. They need reliable retrieval, traceable evidence, permission-aware access and records that can be examined later. By packaging data access and governance with a specialized workspace, OpenAI is trying to address those operational requirements. Independent coverage from Reuters, Business Insider, Folha and Anadolu Agency broadly confirms the product’s target users, design partners and initial data-provider list.
OpenAI plans to expand the breadth of financial information available through the service and extend the product beyond investment banking and equity research. The company also says future model development will focus on identifying, interpreting and applying financial information across tasks normally handled by experienced analysts. No standard public list price was announced in the reviewed sources, so availability will depend on eligibility and enterprise arrangements. Early evaluation should therefore focus on measurable outcomes: the accuracy of retrieved figures, the clarity of citations, the reliability of spreadsheet and presentation outputs, adherence to internal controls and the amount of verified time saved for professionals.
