The Law Society's practice note says solicitors should check client material they know or suspect was AI-generated before relying on it. It also says firms may explain the risks and set preferences for AI use in client communications. These points come from guidance dated 21 October 2026.
Published 8 November 2026. Updated 8 November 2026. Summary of Law Society guidance, not legal advice.
by Margo Legal. Published .
Where does the guidance come from?
The guidance comes from the Law Society's practice note "Using artificial intelligence in legal practice", dated 21 October 2026. Its "Client use of AI" section covers material clients or potential clients produce with AI and send to solicitors. A practice note is guidance, not new legislation.
Most of the note is about solicitors' own use of AI, including AI in existing software.
How might clients use AI before contacting a firm?
The note says firms "may receive AI-generated content from clients or potential clients". It lists five ways clients might use AI systems:
- to develop questions to ask you
- to generate documents, summaries or legal arguments before seeking legal advice
- to draft communications to you
- to summarise communications from you
- to draft complaints about the service they have received
It adds that firms "may encourage clients to disclose any use of AI systems".
What risks does it suggest explaining to clients?
The note warns that AI-generated materials may be inaccurate or lack proper legal reasoning or evidential support. Uploading legal advice without adequate safeguards may undermine confidentiality and privilege. It also warns of additional review costs where clients supply a high quantity of AI-generated material.
It says firms "may wish to cover these risks to clients within your terms of engagement or client care letter". For the confidentiality side more generally, see client confidentiality before AI tools.
What should a firm do with AI-generated material from a client?
Where you know or suspect client material was AI-generated, the note says you should:
- speak with the client to check the suitability and accuracy of any AI-generated materials
- clarify any inaccuracies in the material
- review and validate the legal and factual content before relying on the material
It also says advice should be based on the solicitor's "independent assessment", and that "Professional duties apply regardless of the source of the information or the ways clients develop their communications with you."
Can a firm set rules for how clients use AI in correspondence?
The note says a firm may set out its preferences on AI in client communications. Its example is a request that, if a client uses AI to help draft communications, they disclose that use and the correspondence "remains reasonable and proportionate in length and content". This is separate from the firm's own AI use. The source does not make this example request mandatory.
What about litigants in person?
A separate section covers litigants in person. The note says solicitors may increasingly receive AI-generated or AI-assisted correspondence, complaints, letters of claim or court documents from them, which "may be lengthy, inaccurate or non-compliant with procedural rules". If you think a litigant in person may be using AI to generate court materials, it says you should assess the substance carefully, identify any procedural defects, avoid engaging unnecessarily with irrelevant or repetitive content, and make sure genuine issues are addressed.
Where the material appears to include fabricated authorities, false factual assertions or misleading content, the note says to consider whether and how to raise this with the court. Its considerations include both your duties and the litigant in person's duties to the court. Your client's interests and procedural fairness also matter.
Related reading includes human review of legal AI and telling clients about AI use.
Margo is Margo Legal's product in controlled development. This guidance summary does not depend on using it.
Margo Legal is not a law firm and does not give legal advice. This page summarises the Law Society's practice note dated 21 October 2026.
Frequently asked questions
What does the Law Society say firms should do with AI-generated client material?
Where solicitors know or suspect client material was AI-generated, the note says they should speak with the client to check suitability and accuracy. It also says to clarify inaccuracies and validate the legal and factual content before relying on it.
Must clients disclose every use of AI?
The note says firms may encourage clients to disclose AI use. It also says firms may set preferences for AI in communications, such as requesting disclosure when clients use AI to draft them. It does not present that example request as mandatory.
What risks may firms explain to clients?
The note identifies risks from inaccurate or unsupported AI-generated material. Inputting legal advice without adequate safeguards may undermine confidentiality and privilege. A high quantity of AI-generated material may also take longer to review than expected and add cost.
Should solicitors ignore AI-generated material from litigants in person?
No. The note says solicitors should assess the substance carefully and identify procedural defects. It says to avoid engaging unnecessarily with irrelevant or repetitive content while ensuring genuine issues are addressed. Suspected fabricated authorities or misleading content require consideration of whether and how to raise them with the court.
Sources
- Law Society, Using artificial intelligence in legal practice, 21 October 2026, sections "Client use of AI" and "Litigants in person".