OpenAI launched ChatGPT for Financial Services on September 10, introducing a specialized version of its AI platform tailored for investment bankers and equity researchers, according to Reuters. The product combines the GPT-6 Astra model with built-in financial data integrations from providers like LSEG, PitchBook, and Crunchbase to handle tasks such as financial modeling, research, and pitchbook generation.
Design Partnerships and Industry Focus
The product was shaped by close collaboration with design partners Morgan Stanley and Evercore, as stated in an official OpenAI blog post. According to OpenAI, these institutions helped identify critical workflow friction points, specifically regarding reliable data access and high-quality artifact creation within investment banking and equity research.
The initiative addresses growing demand for generative AI tools within strictly regulated sectors where data governance and security are paramount. ChatGPT for Financial Services incorporates existing enterprise security features, including encryption, role-based access controls, and workspace log exports designed to support internal compliance audit workflows, as reported by Reuters.
Integrated Financial Data Sources
ChatGPT for Financial Services eliminates the need for separate contract negotiations or complex MCP connectors by embedding premium datasets directly into the platform. According to OpenAI, the integrated data includes earnings transcripts, financial statements, and company fundamentals from providers such as:
- Daloopa
- LSEG News
- PitchBook
- Crunchbase
- Quartr
OpenAI hosts and indexes these datasets on its own infrastructure to improve query latency and retrieval accuracy. Furthermore, firms with existing subscriptions can connect proprietary data sources through platform integrations with FactSet, S&P Global, Preqin, and Datasite, according to company statements reported by Reuters.
Granular Citations and GPT-6 Astra Integration
The system relies on GPT-6 Astra, OpenAI’s newest frontier model built specifically to enhance financial reasoning and retrieval precision. OpenAI notes that the architecture allows users to trace generated figures and claims back to specific source tables and passages, highlighting supporting evidence for review during analysis.
For example, when an analyst conducts a P&L normalization analysis, the platform allows them to inspect the reconciliation notes behind an adjusted EBITDA figure to check excluded costs before utilizing the metric in a valuation. OpenAI stated that it plans to continue post-training its models to better interpret financial datasets and eventually expand the product’s scope across the wider financial services sector.
Worth a look