Template

AI policy template for advisers and broker-dealers

Answer six short questions and watch your firm's AI use policy take shape, with a note on why each answer matters. Then complete the parts still in [brackets] and have your CCO or counsel adapt it.

Rule references checked November 2, 2026.

A suggested starting point only, not legal, compliance or regulatory advice. FinSay is not responsible for any policy your firm adopts. Have qualified legal counsel or your chief compliance officer review and adapt it before your firm adopts it. Read the full disclaimer.

Which AI plans are safe for client data? →

Step 1 of 7. Stays in your browser; nothing you enter is sent to FinSay.

Step 1 of 7: Your firm

What is your firm called, and who is your CCO?

How many people work at the firm?

Why this matters

The policy names the CCO as its owner: the person who approves tools, grants exceptions and answers to examiners for it.

Solo practice? Put your own name. The policy still applies to you.

Firm size decides who does what: a solo adviser holds every role, and larger firms also need training tracked. The tool list (section 3) needs one named keeper so it stays current.

Live preview

Showing an SEC-registered adviser until you answer step 2.

[Firm name] — Generative AI Use Policy

Highlighted: you fill this in. Dotted underline: standard wording FinSay suggests; check it fits your firm.

1. Purpose and scope

  • This policy governs how [Firm name] ("the Firm") and everyone who works for it use generative AI tools such as ChatGPT, Microsoft Copilot, Claude and Gemini.
  • It applies to everyone who works for the Firm, including employees, investment adviser representatives, and contractors, on Firm devices and on personal devices used for Firm business.
  • It forms part of the Firm's compliance policies and procedures under Advisers Act Rule 206(4)-7 and is reviewed with them.
  • In this policy, "AI tool" means any generative AI product or feature, including AI features built into software the Firm already uses, such as email, document, CRM and meeting software. "Client information" means any non-public information about a client or prospective client, including the fact that they are one. "AI-assisted" means AI drafted, changed or summarized any part of it.

2. Who is responsible

  • The Chief Compliance Officer, [CCO name], owns this policy. The CCO approves tools and grants exceptions.
  • The CCO, or a person the CCO names in writing, keeps the Approved Tools Register (section 3) current.
  • Each user is responsible for what they enter into an AI tool and for any AI-assisted work they send or rely on.

3. Approved Tools Register

Only tools listed here may be used for Firm business. Approval is for a specific plan, not a brand: a business tier and a personal plan of the same product are different tools.

  • For each tool, record: tool and plan (e.g. "ChatGPT Business", not "ChatGPT"); approved uses; highest data category allowed (section 4); account type users must sign in with; owner; date of last due-diligence review.
  • A tool is added only after the vendor due-diligence review in section 8 and written approval by the CCO.
  • Personal and free plans are not approved for any client information.
  • AI features built into software the Firm already uses (for example Copilot in Outlook or Teams, or AI features in the CRM or meeting software) are separate tools and are not approved just because the software is. Staff do not use one until the CCO approves it, and the CCO turns off unapproved features where the software allows.

4. What data may go into an AI tool

Classify information before entering it.

  • Never, in any tool: Social Security numbers, full account numbers, passwords or login credentials, and government ID numbers such as passport and driver's license numbers. The CCO may add to this list in writing.
  • Client personal information (names, contact details, holdings, financial circumstances): only in tools on the Register approved for client data, signed in with the Firm account, and only what the task needs. Remove anything else.
  • Firm confidential information (strategy, pricing, personnel): only in tools on the Register approved for it.
  • Public information: any approved tool.

5. Permitted and prohibited uses

  • Permitted, subject to this policy: drafting emails and letters, summarizing meeting notes, research on public information, first drafts of educational content, and internal process documentation. Other uses need the CCO's written approval.
  • Prohibited: sending AI-generated content to a client without human review; making or changing investment recommendations based solely on AI output; using AI to generate testimonials, endorsements or performance figures; entering client data into a tool not on the Register; letting an AI tool act on its own, such as sending messages, changing client records, moving money or placing orders, without a person approving each action; and any other use the CCO prohibits in writing.
  • Every recommendation must meet the Firm's fiduciary duty to its clients, whatever part AI played in preparing it.

6. Human review before anything reaches a client

  • A qualified person reviews and edits all AI-assisted client communications before they are sent, and is accountable for them as if they had written them.
  • AI-assisted advertisements and marketing go through the Firm's usual review under the Marketing Rule (Advisers Act Rule 206(4)-1) before use.
  • Users check facts, figures and citations produced by AI against a reliable source. AI output can be wrong while sounding certain.

7. Recordkeeping

  • AI-assisted communications and documents are records like any others and are kept under the Firm's books-and-records procedures (Advisers Act Rule 204-2) for at least five years from the end of the fiscal year in which they were last updated, the first two years in an easily accessible place at the Firm's office, or longer where a rule requires it.
  • Where an AI tool drafts a client communication, the final version sent is retained in [system, e.g. CRM or email archive].
  • Users do not rely on an AI tool's chat history as the Firm's record.

8. Vendor due diligence

  • Before approval, and at least annually after, the Firm reviews each vendor for: whether it trains on Firm data and on which plans; data retention and deletion; security certifications such as SOC 2 Type II; access controls and single sign-on; where data is stored; breach notification terms; and contract terms.
  • Vendors that receive client information are overseen as service providers under the Firm's Regulation S-P procedures, including how quickly they must report a breach to the Firm.
  • Results are recorded with the tool's entry in the Register.
  • When the Firm stops using a tool, the CCO removes it from the Register, ends staff access, has the Firm's data exported where it is a required record, asks the vendor to delete the rest and confirm in writing where it will, and records the date.

9. AI note-takers and call recording

Applies to any tool that records, transcribes or summarizes a meeting or call.

  • Only note-taking and recording tools on the Register (section 3) are used, signed in with the Firm account.
  • At the start of the meeting, before any substantive discussion, the user tells everyone that it will be recorded or transcribed and gets their agreement, noted in the recording or the meeting notes. If anyone declines, the tool is turned off. Some states require every participant's consent; the Firm applies that standard to all meetings.
  • If another participant's AI note-taker joins a Firm meeting, the user asks for it to be removed; if it stays, the user does not discuss client information while it is present.
  • Voiceprint and speaker-identification features stay off unless the Register approves them.
  • The user reviews and corrects an AI-generated summary before it is saved to the client file or sent to anyone. The user is responsible for its accuracy, especially amounts, instructions and anything the client agreed to.
  • The Firm decides in advance which recordings, transcripts and summaries it keeps, where, and for how long, and keeps them under Advisers Act Rule 204-2. The reviewed summary is saved to the client file and kept as in section 7. Raw recordings and transcripts are deleted after 30 days, but only once the CCO has confirmed in writing that recordings of that type are not required records and are not subject to a legal hold or regulatory request.

10. How we describe our use of AI

  • Statements about the Firm's use of AI, in marketing, on the website, in Form ADV and Form CRS, and in conversations with clients, are accurate, specific and supported. The Firm does not claim AI capabilities it does not have or overstate the part AI plays in its advice or recommendations.
  • Written statements go through the review in section 6 before use, and the CCO keeps the support for each one. In conversation, staff describe the Firm's use of AI only as the CCO has approved in writing. A starting description for the CCO to approve: "We use [AI tools the Firm has approved] to help draft and summarize our work. A person at the Firm reviews AI-assisted work before it reaches you. Our professionals, not AI, make every recommendation, and they are responsible for it."
  • Regulators have brought enforcement actions against investment advisers and other firms over misleading claims about their use of AI ("AI washing").

11. Mistakes and incidents

  • Anyone who enters client information into a tool not approved for it reports it to the CCO immediately, and no later than 24 hours after noticing it, without deleting anything first.
  • The CCO assesses it under the Firm's Regulation S-P incident response program, which requires notifying affected clients as soon as practicable and no later than 30 days after learning of certain incidents.
  • The CCO also checks [your state]'s data breach notification law (and that of any state where affected clients live), which sets its own deadlines for notifying clients and, in some cases, the state.
  • Reporting a mistake promptly is expected and will not on its own lead to discipline. Knowingly breaking this policy, or not reporting a mistake, may lead to disciplinary action.

12. Training and attestation

  • Everyone covered by this policy completes AI training on joining and at least annually, and attests in writing that they have read and will follow this policy.
  • The CCO keeps the attestations and training records.

13. Impersonation and deepfakes

AI can imitate a person's voice, face and writing convincingly.

  • Before acting on a request to move money, change account details or share client information, staff confirm it through a separate channel they already trust, such as calling the client back on the number in the Firm's records. A request that arrives by phone, video, email or text is never confirmed through that same call or message.
  • Staff treat urgency, secrecy and last-minute changes to payment instructions as warning signs, and confirm them even when the voice or face seems familiar, including requests that appear to come from colleagues, custodians or vendors.
  • Staff report suspected impersonation or deepfake attempts to the CCO, who alerts the custodian or clearing firm where accounts may be affected, decides what to tell affected clients, and decides whether any filing is required, such as a Suspicious Activity Report for a broker-dealer.
  • Staff do not use AI to create audio, video or images that imitate a real person, including themselves, without the CCO's written approval.

14. Review and exceptions

  • The CCO reviews this policy at least annually, as part of the Rule 206(4)-7 annual review, and whenever an approved vendor changes its data terms or a regulator issues relevant guidance.
  • Exceptions are granted in writing by the CCO, for a stated period, and recorded.
  • Draft prepared [date]. Adopted [adoption date] by [CCO name].

Approval

Adopted by: ____________________ (name and title) Effective date: [adoption date]

Signature: ______________________ Date: __________

[CCO name], Chief Compliance Officer

Before you approve a vendor

Section 8 asks for a due-diligence review. Use the 10-question vendor checklist and compare 22 AI tools' security posture.

Sources

Disclaimer

FinSay LLC ("FinSay") is an educational platform. FinSay is not a law firm, registered investment adviser, broker-dealer or compliance consultant, and this template is not legal, compliance, regulatory, tax or investment advice. Using it does not create an attorney-client, advisory, fiduciary or other professional relationship with FinSay.

This template is a general sample for illustration. It may not include every requirement that applies to your firm, including federal and state law, rules of any self-regulatory organization, your contracts, or the specifics of your business. Filling in the form does not tailor it to your firm or make it advice. Have qualified counsel or your chief compliance officer review and adapt it before adoption. What you type into the form stays in your browser and is not sent to FinSay. To the fullest extent permitted by law, FinSay accepts no responsibility or liability for the content of any policy your firm adopts, or for any loss, regulatory action or other consequence arising from its use.

Laws, regulations, regulatory guidance and vendor terms change. The information here reflects FinSay's review on November 2, 2026, and FinSay does not undertake to update it. Statements about AI products summarize each vendor's public documentation on that date; confirm current terms with the vendor. FinSay is not affiliated with or endorsed by the SEC, FINRA, NASAA or any state securities regulator, or by OpenAI, Microsoft, Anthropic or Google; product names are trademarks of their respective owners.

Your firm remains solely responsible for its own compliance and for any policy it adopts. This content is provided "as is", without warranties of any kind, and your use of it is governed by the FinSay Terms of Service, including their disclaimers and limitation of liability.