OpenAI has announced ChatGPT for Financial Services, a new platform for the financial sector that will be used for research, financial modeling and the preparation of client materials.
According to the company, the product combines GPT-6 Astra’s reasoning capabilities with financial data and tools. Its development benefited from design partnerships with the US-based investment banks Morgan Stanley and Evercore.
Integration with financial data providers
The platform will work with datasets from providers such as Daloopa, PitchBook and LSEG News. These datasets will include company financial statements, earnings-call transcripts, company fundamentals and information on private companies.
OpenAI said the data would be indexed and hosted on its infrastructure. The company also said the platform would provide detailed citations, allowing users to trace financial figures and claims back to their sources.
ChatGPT for Financial Services is also planned to integrate with existing subscription services offered by S&P Capital IQ, LSEG, MSCI, Dow Jones Factiva and Moody’s. Through a single sign-on system, financial professionals will be able to access data to which their institutions have access rights.
More than 50 integrations and document analysis
OpenAI said it was expanding integrations with financial data providers and had built an ecosystem of more than 50 integrations, including Datasite, Box, Preqin and Intapp.
According to the company, GPT-6 Astra can interpret figures, tables and supporting notes in financial documents. The system can also conduct financial analysis and prepare documents, spreadsheets and presentations.
OpenAI said GPT-6 Astra scored 69.9 percent on the OfficeQA Pro test, which measures the ability to find and analyze information in US Treasury documents, while GPT-5. 6 Sol scored 60.2 percent.
Internal oversight and security tools
Administrators will be able to publish approved Excel, Word and PowerPoint templates through the platform. Teams will be able to use these templates to prepare valuation models, research notes and presentation files that comply with their institutions’ formatting and style rules.
The product also includes ChatGPT Enterprise security and governance features such as single sign-on, role-based access controls and configurable data-retention policies.
OpenAI said that, by default, commercial data from financial institutions would not be used to train its models and that the data would be encrypted in transit and at rest. Compliance teams will be able to export supported workspace logs for audits and investigations, while institutions will be able to control employees’ access to specific skills, applications and data sources.
ChatGPT for Financial Services is now available to eligible financial institutions.
