OpenAI today unveiled ChatGPT for Financial Services, a tailored environment that combines hosted financial datasets with the company’s GPT‑6 Astra model to support research, financial modeling and client‑facing deliverables. ChatGPT for Financial Services is positioned to bring the latest frontier model capabilities into workflows used by investment bankers, equity researchers and other finance teams, with an emphasis on information retrieval, financial reasoning and artifact generation.
At the heart of the offering is GPT‑6 Astra, which OpenAI describes as optimized for navigating tables and notes in financial documents, running analyses and synthesizing results into documents, spreadsheets and slides. OpenAI cited internal benchmarks showing GPT‑6 Astra scored 69.9% on the OfficeQA Pro test, compared with 60.2% for GPT‑5.6 Sol, underscoring the company’s claim of improved performance on office and document tasks relevant to financial professionals.
A distinguishing element of the product is its integrated, premium financial data hosted on OpenAI’s infrastructure. At launch, OpenAI includes indexed datasets from Daloopa, PitchBook, LSEG News and Crunchbase, covering earnings transcripts, financial statements, company fundamentals and private company information. By hosting and indexing these sources, OpenAI says firms can avoid separate connector setup or new contracts for those included datasets and can use granular citations that let users trace figures and claims back to original tables and passages.
OpenAI is also pursuing integrations that let firms access data tied to their existing subscriptions. The company is working with providers including S&P Capital IQ, LSEG, MSCI, Dow Jones Factiva and Moody’s on shared sign‑in and entitlement recognition so access can be granted based on a user’s ChatGPT sign‑in. For commonly used connections, OpenAI has optimized a selection of MCP connectors and supports a broader connector ecosystem of more than 50 sources, among them Datasite, Box, Preqin, FactSet and Intapp.
The product is built to convert analysis into firm‑ready artifacts. Administrators can publish Excel, Word and PowerPoint templates through an admin interface so teams can automatically generate valuation models, research notes and pitchbooks in approved formats and styles. OpenAI highlighted the ability to create interactive charts and visualizations with underlying source data available for review, and to produce deliverables that adhere to a firm’s formatting and compliance requirements.
Design partnerships with Morgan Stanley and Evercore helped shape the initial focus on investment banking and equity research. OpenAI said feedback from those collaborations pinpointed reliable access to data and high‑quality artifact creation as primary pain points, which guided the product’s early feature set and future training priorities.
Enterprise security and governance are central to the offering. ChatGPT for Financial Services builds on ChatGPT Enterprise controls such as SAML single sign‑on, SCIM provisioning and role‑based access controls. OpenAI stated that firms’ business data is not used to train its models by default, and that data is encrypted both in transit and at rest. Administrators can configure workspace retention policies, and compliance teams can export supported workspace logs through the OpenAI Compliance Platform into existing audit workflows. The product also supports role‑based management of skills and apps and the creation of multiple workspaces to maintain information barriers.
OpenAI said GPT‑6 Astra will be available natively within ChatGPT for Financial Services and that newer models will be added as they are released. Financial institutions and developers may also use OpenAI’s API to build specialized applications that combine frontier models with domain expertise, data providers and software partners, enabling bespoke tools that leverage the same core capabilities.
ChatGPT for Financial Services is available to eligible financial institutions. Organizations interested in deploying the offering are asked to contact OpenAI or their account teams for more information. The product aims to deliver a managed environment where finance teams can accelerate research, modeling and client‑facing work while retaining enterprise controls and traceability.
Source: Read the original source
Leave a Reply