OpenAI has launched a separate ChatGPT plan for eligible financial institutions, combining GPT‑6 Astra with included financial datasets, entitlement-linked sources and firm templates. Pricing remains private and access is not self-service.
RELATED BUYER PROFILEReview access, pricing evidence and the test-pending product profilePricing · access · strengths · limitations →CONTINUE THE DECISIONBuild a controlled AI stack for a finance teamEvidence · workflow · next action →CONTINUE THE DECISIONReview product, dataset and governance evidenceEvidence · workflow · next action →CONTINUE THE DECISIONCalculate complete cost per accepted finance deliverableEvidence · workflow · next action →What you need to know
- ChatGPT for Financial Services is a separate plan built on ChatGPT Enterprise, initially focused on investment banking and equity research
- Some datasets are included while other sources require existing subscriptions, entitlements and workspace permissions; coverage and delays vary
- OpenAI has not published a standard price or self-service trial, and standard Enterprise and Financial Services seats cannot be mixed in one workspace
The procurement boundary
What OpenAI launched
ChatGPT for Financial Services is a separate ChatGPT plan for eligible financial institutions, built on ChatGPT Enterprise. OpenAI says the initial product is designed for investment banking and equity research: company research, earnings analysis, peer comparison, valuation work, editable financial models, research notes and pitchbooks. It combines GPT‑6 Astra with financial data and firm-controlled templates rather than selling a standalone model or a general retail-investing assistant.
Access and pricing are controlled
There is no public list price or self-service trial. Organisations must contact OpenAI Sales or their account team, and eligibility depends on the organisation's size and type as well as applicable data-provider restrictions. OpenAI's help guide also states that the plan applies to the entire workspace: standard ChatGPT Enterprise and Financial Services seats cannot be mixed in that workspace. Buyers should obtain a written quote covering seats, included data, usage, implementation, support and any connected-provider cost before comparing alternatives.
Included data does not mean complete or real-time data
OpenAI lists included access to public-company filings, Quartr, Daloopa, Fiscal.ai, PitchBook Essentials, Crunchbase, LSEG News for US-based financial professionals, and US equity pricing from FMP and Nasdaq. Other providers can recognise existing customer entitlements through connected access. The boundaries matter: included coverage may be narrower than the provider's full product, Daloopa is listed with a 24-hour delay and 3,000 datapoints per user per month, and Nasdaq data supplied through FMP is delayed by 15 minutes. Availability can change, so the source, timestamp, geography and entitlement should travel with every material figure.
Citations help—but reconciliation is still required
The product can trace figures and claims to tables and passages, then turn analysis into editable Excel, Word and PowerPoint materials using firm templates. That is useful evidence infrastructure, not an accuracy guarantee. Analysts should inspect the cited row and period, rebuild consequential calculations outside the generated narrative, confirm that adjustments are intentional and preserve reviewer sign-off before a result reaches a client or investment decision.
Governance must cover both firm data and licensed data
OpenAI states that the plan inherits ChatGPT Enterprise controls including SAML SSO, SCIM, role-based access, configurable retention and supported audit exports, and that business data is not used for model training by default. Those statements do not remove the institution's own obligations. A UK buyer should verify the exact contracted processing and residency terms, information barriers, role permissions, retention, source licences, redistribution limits and whether connected data can appear in client materials. Vendor statements are inputs to due diligence, not a HubAI compliance badge.
HubAI buyer verdict
ChatGPT for Financial Services is a credible shortlist candidate for institutions that already buy multiple research datasets and spend significant analyst time moving from source retrieval to models and client-ready materials. The strongest pilot is one real recurring workflow with an approved baseline: compare coverage, citation validity, calculation accuracy, correction time, permissions, output quality and complete cost. Do not use it as financial or investment advice, and do not adopt it solely because GPT‑6 Astra or a familiar data-provider name is included.
HUBAI VIEWFinance buyers should test one reconciled deliverable end to end—data coverage, citations, calculations, permissions, review time and complete cost—not treat an integrated data list as proof of accuracy.
Buyer decision signal: New finance plan · controlled enterprise access
What to verify next
1Confirm organisation eligibility, UK access and the exact workspace-wide commercial quote
2List which datasets are included, entitlement-linked or unavailable for each user and jurisdiction
3Record coverage, timestamp, delay, usage limit and redistribution rules for every consequential source
4Reconcile figures, formulas and adjustments to an analyst-approved baseline before client or investment use
5Test information barriers, role permissions, retention, audit exports and account offboarding
6Keep source citations and qualified reviewer approval in every shared deliverable
7Measure correction time, failed retrievals, stale-data incidents and complete cost per accepted output
8Compare the pilot with the current data, spreadsheet, research and presentation stack
Read the evidence
Capabilities, availability and prices can change. HubAI keeps analysis separate from the underlying official material.
01OpenAI: Introducing ChatGPT for Financial ServicesOpen source ↗02OpenAI Help: access, included datasets and workspace controlsOpen source ↗03OpenAI: business data privacy, security and governanceOpen source ↗
Models
Safety
Products
Products