Skip to main content

OpenAI's New Finance Chatbot: GPT-6 Astra Takes on Wall Street

OpenAI's New Finance Chatbot: GPT-6 Astra Takes on Wall Street

OpenAI is making a bold move into investment banking. On September 10, the company unveiled ChatGPT for Financial Services, a customized version built with Morgan Stanley and Evercore. The product initially targets investment banking and stock research, signaling that large language models are now ready for the core operations of financial institutions.

What It Can Do

At its heart, the service runs on GPT-6 Astra, OpenAI's latest inference engine. That means it can efficiently process numbers, tables, and footnotes buried in massive financial documents. Financial teams can research across multiple data sources, track data changes over time, generate interactive charts, and even build complex financial models. They can also write research reports and create customized client pitch materials—all within the tool.

Data and Compliance

Accuracy and compliance are non-negotiable in finance. To that end, the product includes authoritative data from providers like Daloopa, PitchBook, LSEG News, and Crunchbase. This covers earnings calls, financial statements, company fundamentals, and private company information. All data is indexed and hosted by OpenAI, so users can make fine-grained citations and trace any figure back to its source.

OpenAI also plans to work with S&P Global, LSEG, MSCI, Dow Jones Factiva, and Moody's, letting users access licensed content through their existing subscriptions. The company has optimized MCP connectors for financial data providers, integrating more than 50 connectors including Datasite, Box, Preqin, and Intapp.

Built for the Enterprise

Financial institutions can publish their own templates—Excel, Word, PowerPoint—through a dedicated admin page. Once administrators set up templates and format specs, team members can quickly turn analysis into valuation models and professional documents that meet institutional standards. Developers can also build custom applications via API.

On the security front, the product inherits ChatGPT Enterprise standards: SAML single sign-on, SCIM user provisioning, and role-based access control. OpenAI promises that enterprise data won't be used for model training by default, and data is encrypted both in transit and at rest. Admins can set data retention periods, and compliance teams can export workspace logs via the OpenAI Compliance Platform for audits. Enterprises can also manage read/write permissions by role and enforce information isolation through multiple workspaces.

The product is now available to qualified financial institutions. For an industry that thrives on information advantage, this could be a game-changer—or at least a powerful new tool in the arsenal.

Key Points

  • OpenAI launches ChatGPT for Financial Services, developed with Morgan Stanley and Evercore, targeting investment banking and stock research.
  • Powered by GPT-6 Astra, it handles numbers, tables, and footnotes in complex financial documents, enabling modeling, research reports, and client pitches.
  • Integrated data from Daloopa, PitchBook, LSEG News, Crunchbase, and more, with plans to add S&P Global, MSCI, and others.
  • Enterprise-grade security: SAML SSO, SCIM, role-based access, encrypted data, and no training on enterprise data by default.
  • Available now to qualified financial institutions.