OpenAI has launched ChatGPT for Financial Services, a specialised version of its AI platform built for investment bankers, equity researchers and other professionals working with complex financial data.
The product combines OpenAI’s latest GPT-6 Astra model with datasets from major financial information providers. Morgan Stanley and Evercore worked with OpenAI as design partners while the system was being developed.
The launch pushes ChatGPT deeper into one of the world’s most heavily regulated industries.
Instead of simply asking an AI model to explain a company, users will be able to combine institutional financial data, internal company information and AI reasoning in a single workflow.
Proposal for Conducting Cyber Crisis Drill, Tabletop Exercise (TTEx) & CCMP Readiness Exercise
From earnings transcripts to pitchbooks inside one AI system
ChatGPT for Financial Services can search and analyse data covering company fundamentals, earnings transcripts, financial statements and market information.
OpenAI has built in datasets from providers including Daloopa, LSEG News, PitchBook, Crunchbase and Quartr. The company says those datasets are indexed on its own infrastructure to make retrieval and citations more reliable.
Firms that already pay for specialist financial databases do not necessarily have to abandon them.
OpenAI says organisations can connect existing subscriptions from providers including FactSet, S&P Global, Preqin and Datasite. Its broader financial-services offering also supports bringing market, company and internal data into connected AI workflows.
For an investment banker, that could mean asking ChatGPT to compare companies, analyse financial statements, help build a valuation model and then prepare portions of a client presentation using the bank’s own templates.
Equity researchers could use it to combine earnings transcripts, historical financial data and company filings before preparing research.
Those tasks traditionally require analysts to move repeatedly between spreadsheets, financial databases, company documents and presentation software.
OpenAI is effectively trying to place an AI reasoning layer across those separate systems.
What makes financial AI different from an ordinary chatbot?
Financial work creates a problem that ordinary consumer AI does not always face: a plausible-sounding mistake can become extremely expensive.
An incorrect figure in a valuation model, a missed disclosure or a fabricated source could affect an investment decision, transaction or client recommendation.
That is why traceability matters.
In simple terms, a financial professional should be able to see where an AI obtained a number or statement rather than trusting an answer simply because it sounds confident.
OpenAI says its financial workflows are designed around structured, cited outputs and connections to approved data sources. Its existing financial-services tools are also built to work with spreadsheets, filings, transcripts and proprietary company information.
GPT-6 Astra has been tuned to improve financial reasoning, retrieval across data tools and the accuracy of generated material, according to OpenAI.
But human review remains critical.
OpenAI’s own finance material says sensitive decisions should remain subject to human approval, particularly when AI is working with forecasts, capital allocation, compliance or other high-impact financial processes.
Security and compliance are central to the pitch
Banks and investment firms cannot simply upload confidential deal information into an uncontrolled AI system.
They may hold unpublished earnings information, merger documents, client portfolios, trading strategies and personally identifiable information.
OpenAI therefore says the new product builds on ChatGPT Enterprise security controls.
These include encryption, role-based access and the ability for compliance teams to export workspace logs into existing audit processes.
Role-based access means not every employee automatically sees every dataset.
A junior analyst, compliance officer and senior banker can be given different permissions depending on what they are authorised to access.
Audit logs provide another safeguard. They allow an organisation to retain records of how the system was used and potentially reconstruct what happened if an output later creates a compliance concern.
OpenAI’s broader financial-services platform also emphasises configurable data controls and keeping enterprise information under the customer’s control.
These safeguards will be particularly important because financial institutions operate under strict rules around confidentiality, record-keeping and supervision.
Morgan Stanley has already been testing AI at scale
OpenAI’s relationship with Morgan Stanley predates the new product.
The two companies previously worked together on AI systems for Morgan Stanley financial advisers. OpenAI says the bank built an evaluation framework to measure whether AI produced sufficiently reliable and consistent answers before expanding its use.
OpenAI has said that 98% of Morgan Stanley advisers eventually adopted its AI tools, while access to internal documents rose significantly and information-retrieval times fell.
The new launch takes that idea further.
Rather than developing separate AI tools for individual institutions, OpenAI is now offering a financial-services platform that can connect directly with datasets and workflows used across the industry.
It also reflects a wider change in the AI business.
OpenAI is increasingly building specialised products for industries where companies require more than a general-purpose chatbot. The company has been expanding into areas such as financial services, chip design and life sciences as enterprise customers look for AI that can deliver measurable operational value.
OpenAI says it intends to expand ChatGPT for Financial Services beyond investment banking and equity research into other parts of the industry.
That could eventually include more workflows across asset management, banking, risk, compliance and other financial functions.
What this means for you: AI is moving from answering financial questions to working directly inside the systems professionals use to analyse companies and make decisions. For customers and investors, the key issue will be whether financial institutions maintain human review, traceable sources and strict controls over sensitive data.
The420 Insight: The competition in enterprise AI is shifting from who has the smartest general chatbot to who can embed AI inside specialised, regulated workflows. In finance, access to trusted data, auditability and compliance may ultimately matter as much as raw model intelligence.
Follow for daily updates on cybercrime, corporate fraud, DFIR, hacking, investigations, and digital forensics
