ChatGPT for financial advisors: what the SEC and FINRA rules mean, and what your firm’s AI policy should say

By Dan Blomquist, Aegis AI Advisory · Published November 17, 2026 · Sources checked November 17, 2026

Yes, financial advisors can use ChatGPT, and no SEC or FINRA rule bans it. There is no AI exemption either. FINRA says its rules are technology neutral and apply to generative AI the same way they apply to any other tool, and the SEC has fined two advisers for overstating how they used AI. For a small advisory firm, that comes down to four habits:

1. Client account data stays out of personal AI accounts.
Statements, tax returns, Social Security and account numbers go only into tools your firm has approved after reading the vendor’s terms.
2. AI drafts are your communications.
A client email, a newsletter or a website chatbot answer is reviewed and kept like anything else your firm sends. That is true whether a person or a model wrote it.
3. Your AI claims have to be true.
“AI-driven planning” on your website or in your brochure is a marketing claim. Examiners say they will check it.
4. A person owns every recommendation.
AI can summarize, draft and organize. The advisor, who owes the client a fiduciary duty or a best-interest obligation, decides.

Search for “ChatGPT for financial advisors” and you mostly find prompt lists. This page covers what those lists leave out: which rules apply to your firm, what is fine to put into an AI tool, and the short written policy that ties it together. For the general structure of any staff AI policy, start with what an AI acceptable-use policy should contain.

Which rules reach your firm depends on how you are registered

“Financial advisor” covers several kinds of firms, and the rules differ. The SEC says advisers with more than $110 million in regulatory assets under management generally must register with the SEC, while those under $100 million generally may not and register with their state instead. Many small practices are in that second group.

Your firmRules that matter most for AI use
SEC-registered investment adviserThe Marketing Rule (which applied in the 2024 AI-washing cases); books and records under Rule 204-2, including written communications relating to advice; Regulation S-P as amended in 2024 (incident response, service provider oversight, customer notice); and the SEC’s FY 2026 exam priorities, which name AI representations and AI supervision.
State-registered adviser (in Michigan, under the Michigan Uniform Securities Act)Your state’s adviser rules. Michigan’s R 451.4.21 requires written physical security and cybersecurity policies and an annual privacy policy for clients. Regulation S-P’s list of covered institutions names advisers registered with the SEC, so state-registered firms should look first to the state rules.
Broker-dealer or dual registrant (FINRA member)FINRA Regulatory Notice 24-09; Rule 3110 supervision; Rule 2210 content standards, which FINRA says apply to AI-created communications and to investor-facing chatbots; Regulation Best Interest if AI is used to make a recommendation to a retail customer; and Regulation S-P.

There is no SEC “AI rule,” and that is the point

In 2023 the SEC proposed rules on conflicts of interest from “predictive data analytics” used by broker-dealers and advisers. On June 12, 2025 it formally withdrew that proposal, saying it would publish a new proposal if it decides to act in that area again. So when someone tells you the AI rules for advisers are “coming,” the honest answer is that the rules that apply today are the ones you already have: fiduciary duty, marketing, recordkeeping, privacy and supervision.

Regulators have said how they read those rules for AI. The SEC’s Division of Examinations says in its fiscal year 2026 priorities that it will review registrants’ representations about their AI capabilities for accuracy and assess whether firms have adequate policies and procedures to monitor or supervise their use of AI. FINRA’s 2026 oversight report added a GenAI section. It reports that the top use among member firms is summarization and information extraction, and it suggests firms consider:

A one-person RIA does not need a model risk team. It does need to be able to say which tools it uses, for what, and who checks the output.

Before you paste: a quick sort of common advisory tasks

This assumes a personal or free AI account unless your firm has approved a business plan.

TaskClient data?What to do
Summarize a fund prospectus, a 10-K or an earnings call transcriptNoFine. Check any number you reuse against the source; FINRA lists hallucination as a core GenAI risk.
Draft a market commentary, newsletter or social postNoFine with review. It is still your firm’s communication or advertisement, so the usual content and performance-claim rules apply before it goes out.
Paste a client’s statement, tax return or account numbers into a chatbotYesNot in a personal account. The SEC lists a Social Security number, or an account number combined with information that could be used to access the account, as examples of sensitive customer information. Use only an approved tool whose terms you have reviewed.
Draft a follow-up email that summarizes your recommendationUsuallyCareful. You review it before it is sent. Once sent, it is a written communication relating to advice that Rule 204-2(a)(7) requires an SEC-registered adviser to keep, so send it from a system that archives it.
AI notetaker on client meetingsYesCareful. The vendor holds client information, so treat it as a service provider you oversee. Decide where the notes live, and whether any summary sent to the client is being retained. Recording-consent rules vary by state.
Website chatbot that answers investor questionsOftenCareful. FINRA’s FAQ B.4 says chatbot communications may be correspondence, retail or institutional communications and must be supervised. Our look at whether an AI receptionist is worth it covers the operating side.
Describe your process as “AI-powered” on your site or in Form ADVNoOnly if it is true and you can show how. This is what the SEC charged Delphia and Global Predictions over.
Let an AI tool pick investments or rebalance a client account without reviewYesNo. A person makes and documents the recommendation. FINRA’s FAQ notes that Regulation Best Interest applies when a broker-dealer uses AI to make a recommendation to a retail customer.

A starting point, not legal advice. Your broker-dealer, custodian, compliance consultant or state regulator may require more.

Why the account type matters

OpenAI’s help center says that for its services for individuals (Free, Plus and Pro) it may use your content to train its models unless you turn off “Improve the model for everyone,” and that giving a thumbs-up or thumbs-down can send the whole conversation into training even after you opt out. For ChatGPT Business, Enterprise and the API, OpenAI says it does not train on inputs or outputs by default, and its enterprise privacy page describes admin-controlled retention and a data processing addendum for those plans.

That difference matters under Regulation S-P. The 2024 amendments require covered firms to have written policies for overseeing service providers, including making sure the provider protects customer information and notifies the firm within 72 hours of certain breaches. A business AI plan is a vendor relationship you can review against that list. An advisor’s personal chatbot login is not a vendor relationship at all. The same question applies to Copilot, Gemini and the AI features now built into CRM, planning and archiving software: which plan is your staff actually signed into, and what do its terms say?

Notes for Michigan advisers

What goes in an advisory firm’s AI policy

The general sections follow the same pattern as any firm’s policy; compare the free sample policy. Firms that also prepare tax returns will want the IRC 7216 rules in our CPA firm guide, and advisers who also sell insurance can compare the insurance agency guide.

Sample clauses for an advisory firm

Excerpt for illustration only. Not legal advice. [Firm] and [CCO] are placeholders.

2. Approved tools

Only AI tools and plans on [Firm]’s Approved AI Tools list may be used for firm business. The list records, for each tool, whether it may receive client information and whether its outputs are captured by the firm’s archiving system. [CCO] maintains the list and reviews each vendor’s data terms before approval and at least annually.

3. Client information

No one may enter a client’s nonpublic personal information, including Social Security numbers, account numbers, statements, tax returns or financial plans, into an AI tool that is not approved for client information. Personal or free AI accounts may never be used for client work.

4. Advice and review

AI output is a draft. An investment adviser representative must review any AI-assisted recommendation, plan summary, performance discussion or client message before it is used, and is responsible for its accuracy and suitability. No AI tool may place, change or recommend a transaction for a client without that review.

5. Communications, marketing and records

AI-assisted client communications and advertisements follow the same review and retention procedures as any other firm communication. The firm will not describe its use of AI in any advertisement, brochure or website unless [CCO] confirms the description is accurate and documented.

7. Mistakes

If client information is entered into a tool that is not approved, tell [CCO] the same day. [CCO] will decide whether it is an incident that requires notice under the firm’s incident response procedures.

A 30-minute self-check

If you answer “no” or “not sure” to any of these, that is where your policy should start.

QuestionYes / No
Do you know every AI tool your staff use for firm work, including ones built into your CRM or email? Our shadow AI guide explains why this comes first.☐
For each one, do you know the plan and whether it trains on your data?☐
Are AI-drafted client emails sent from a system that archives them?☐
Does any AI notetaker or chatbot vendor appear on your service provider list?☐
Is every statement about AI on your website and in your brochure accurate today?☐
Does your incident plan say what happens if client data goes into the wrong AI tool?☐

Then write it down and get signed acknowledgments. The 90-day AI governance plan shows how to sequence the rest.

What isn’t settled

The SEC withdrew its predictive analytics proposal but kept the door open to a new one, and exam priorities change every year. FINRA’s notice says it may issue more guidance on specific uses. It is not always clear when an AI meeting summary or an internal draft becomes a record you must keep, and firms and their compliance consultants draw that line differently. State rules for advisers vary. Vendors also change which plans train on your data, so the account-type section above is a snapshot as of November 17, 2026. A written policy makes your choices explicit and easy to update. It does not replace your compliance consultant, your broker-dealer’s requirements or an attorney.

Get a policy drafted for your firm

Answer 10 questions, choose Financial advisory, and the generator emails you an AI acceptable-use policy that addresses SEC Regulation S-P, SEC and FINRA recordkeeping and communications rules (including AI-drafted client communications), fiduciary duty and suitability review of AI output, using the tools and state you enter.

Build my firm’s AI policy →

Not ready? Read the free sample policy first, or fill in our $9 DIY Starter Kit.

Disclosure: Aegis AI Advisory sells this generator. It produces a template, not legal advice, and it does not review your compliance manual, vendor contracts or Form ADV. It is not the right fit if your broker-dealer or compliance consultant has already given you an AI policy you must follow, or if you need written supervisory procedures for AI-driven trading or advice tools. For a firm that wants an AI usage inventory, vendor review and staff training, see the AI governance track or book a free AI Operations Briefing.

Frequently asked questions

Can financial advisors use ChatGPT?

Yes. No SEC or FINRA rule bans it, and FINRA says its rules are technology neutral: they apply when a firm uses generative AI just as they apply to any other tool. The limits come from rules you already follow. Keep client nonpublic information, such as Social Security and account numbers, out of personal Free, Plus or Pro accounts, where OpenAI says it may use conversations to train its models unless the user opts out. Treat anything AI drafts for a client or the public as your firm's own communication, reviewed and kept like any other. And do not describe your firm's AI use in a way you cannot back up.

Is there an SEC rule on AI for investment advisers?

Not an AI-specific final rule. The SEC proposed rules on conflicts of interest from predictive data analytics in 2023, then formally withdrew that proposal on June 12, 2025, saying it would issue a new proposal if it decides to act in that area again. Existing rules still apply, and the SEC's Division of Examinations says in its fiscal year 2026 priorities that it will review the accuracy of firms' representations about their AI capabilities and whether firms have adequate policies and procedures to monitor or supervise their use of AI.

What does FINRA say about generative AI?

FINRA Regulatory Notice 24-09 (June 27, 2024) reminds member firms that FINRA rules and the securities laws continue to apply when they use generative AI and large language models, including tools embedded in third-party products. It points to Rule 3110 (Supervision) and Rule 2210 (Communications with the Public) as examples. FINRA's 2026 Annual Regulatory Oversight Report added a GenAI section that suggests firms consider review and approval processes for AI tools, testing, and ongoing monitoring of prompts and outputs, which may include keeping prompt and output logs and tracking which model version was used.

Are AI-written client emails and marketing subject to recordkeeping?

For FINRA member firms, FINRA's advertising FAQ D.8 says firms are responsible for their communications whether a human or AI wrote them, including supervision, applicable FINRA and SEC recordkeeping requirements and the content standards in Rules 2210 and 2220. For SEC-registered investment advisers, Rule 204-2(a)(7) requires keeping written communications sent and received relating to any recommendation or advice. Nothing in that rule exempts a message because AI drafted it.

What is AI washing?

It is claiming to use AI in ways you do not. On March 18, 2024, the SEC announced settled charges against two investment advisers, Delphia (USA) Inc. and Global Predictions Inc., for false and misleading statements about their use of AI; they agreed to pay $400,000 in total civil penalties ($225,000 and $175,000). Both were also charged under the Marketing Rule. A firm website or brochure that calls your process AI-driven needs to be accurate.

Does Regulation S-P apply to the AI tools my firm uses?

If your firm is a covered institution, such as a broker-dealer or an investment adviser registered with the SEC, the 2024 amendments require a written incident response program; notice, with limited exceptions, to individuals whose sensitive customer information was or is reasonably likely to have been accessed without authorization, as soon as practicable and no later than 30 days after becoming aware of the incident; and policies requiring oversight of service providers, including that they notify you within 72 hours of certain breaches. An AI vendor that stores your clients' information is a service provider for this purpose. Larger entities had to comply by December 3, 2025 and smaller entities by June 3, 2026.

Does a state-registered adviser in Michigan need an AI policy?

No Michigan rule uses the words AI policy. But Michigan's investment adviser rule, R 451.4.21, requires a registered adviser to keep written physical security and cybersecurity policies that safeguard confidential client records, and to deliver a privacy policy at engagement and annually. An AI tool that holds client records falls inside those policies, so writing down which tools are approved for client data is the practical way to show you meet them.

What does the Aegis AI policy generator produce for an advisory firm?

Choose Financial advisory as your industry and answer 10 questions. The $49 Policy is an AI acceptable-use policy in an editable Word file and a web version, plus a one-page staff checklist. For advisory firms it is written to address SEC Regulation S-P, SEC and FINRA recordkeeping and communications rules (including AI-drafted client communications), fiduciary duty, and suitability review of AI output, using the tools and state you enter. The $149 Pro Pack adds an AI vendor risk register and a staff training sign-off sheet. It is a template, not legal advice.

Sources

Also free: check whether ChatGPT and Google’s AI can find your firm → · More guides on the resources page.