Skip to main content
Security and ethics

Firm AI policy template for lawyers (fill-in outline)

A law firm AI policy should say which tools are approved, what information may never be entered, how output is verified by a lawyer, when clients are told, and who owns the policy. This page gives a fill-in outline you can adapt, written as general guidance and not legal advice.

By Raghav Handa · Updated 17 October 2026 · 3 min read

How to use this outline

Copy the headings into your own document and replace the bracketed prompts. Treat the wording as a starting point, have it reviewed by whoever advises your firm on professional-conduct matters, and align it with your insurer's requirements and your regulator's guidance.

General information only, not legal advice. Check the rules of your own bar or regulator and your clients' terms.

Policy outline

1. Purpose and scope

  • [Firm name] permits the use of AI tools to support, not replace, professional judgment.
  • Applies to: [partners, associates, paralegals, contractors, interns].
  • Covers: [generative AI chat, drafting, summarisation, transcription, research, AI features inside other software].

2. Approved tools

  • Approved list: [tool, plan or tier, approved use, approver, date].
  • Unapproved tools and personal accounts must not be used with client information.
  • Requests for new tools go to: [name or role].

3. Confidential information

  • Never enter: [client-identifying details, privileged advice, personal data, trade secrets] into [unapproved tools].
  • Where anonymisation is required, describe the standard: [remove names, dates, identifiers].
  • Where the tool is approved for client data, record the settings required: [training off, retention, access].

4. Verification and competence

  • A lawyer reviews all AI output before it is relied on or sent.
  • Every citation and quotation is checked against a primary source.
  • The responsible lawyer remains accountable for the work product.

5. Client communication

  • When disclosure or consent is needed: [criteria].
  • Template used: see the consent template page.
  • Fee treatment of AI-assisted work: [approach].

6. Supervision, training and incidents

  • Training: [on joining and annually].
  • Report suspected misuse or leaks to [role] within [timeframe].
  • Consequences of breaching the policy: [text].

7. Review

  • Owner: [name]. Review date: [date]. Version log: [table].

Keeping the register in Notion

A simple Notion database can hold the approved-tools register: tool, owner, approved use, data allowed, review date. That gives you one place to check and a record of when each tool was last reviewed. LawyerOS does not include an AI policy or compliance register by default; this is something you add yourself.

Frequently asked questions

Does every firm need an AI policy?
Regulators and professional bodies increasingly expect firms to have some governance around AI use. Even a one-page policy helps; check your own regulator's expectations.
Who should own the AI policy?
Usually a partner or the person responsible for compliance or risk, with input from IT and practice leads.
Should the policy ban public AI chat tools?
Many firms prohibit client information in consumer tools while allowing approved enterprise versions. The right line depends on your risk appetite and clients' terms.
Is this a legal document?
No. It is a general outline. Adapt it and have it reviewed for your jurisdiction.

About the author

Raghav Handa

Raghav Handa trained as a lawyer, then spent years building legal technology, from enterprise legal products to custom systems for law firms.

Want it built for you?

LawyerOS is a ready-made Notion workspace for matters, deadlines, time and invoices. Starter $19, Pro with AI $39, one-time.

See LawyerOS