What is CILEx Regulation's AI guidance and who does it cover?
CILEx Regulation published its Guidance on the Use of Artificial Intelligence on 1 October 2026. In its own words, it "explains how the existing requirements of the CILEx Code of Conduct apply when AI is used". It "applies to all individuals and entities regulated by CILEx Regulation" (paragraph 4).
Regulated individuals working in organisations regulated by the SRA or another regulator "should also have regard to any relevant requirements and guidance that apply to them in that context". For the SRA's own position, see our note on the SRA's AI warning notice.
How binding is the guidance?
The guidance is framed as explaining existing duties rather than adding new ones. Paragraph 2 says "The use of AI does not alter your professional responsibilities", and the document links its points back to the existing Code of Conduct.
Its wording varies in strength. Some points use "must", some use "should", and some are framed as options a firm "may also wish" to take up. CILEx Regulation also says it will review the guidance periodically (paragraph 3).
Who is responsible for AI-assisted work?
Regulated individuals and entities remain accountable, whether or not AI was used (paragraph 8). Paragraph 10 says "You cannot delegate responsibility for your work to the AI tool or its provider." Professionals "should be able to explain the basis on which AI-assisted work has been reviewed, amended and approved before it is used".
Our piece on what human review means for legal AI looks at that kind of checking more generally.
Which uses does it treat as higher risk?
Paragraph 6 separates lower-risk tasks, such as administrative support and summarising documents, from higher-risk uses. Those "may include generating substantive legal analysis, advice, pleadings, witness evidence or material intended for courts, tribunals or regulators", and the guidance says they "require closer scrutiny, verification and oversight".
Does it say clients should be told about AI use?
Paragraph 19 says regulated individuals and entities "should consider whether clients ought to be informed". Disclosure "is likely to be appropriate" where AI "materially contributes to the service provided", where confidential or personal information goes into an AI system, or where a client might otherwise be misled about the extent of human involvement.
Any explanation "should be clear, accurate and not misleading". We cover this more broadly in telling clients about AI use, and the confidentiality side separately.
Does a firm need an AI policy or an AI Statement?
Paragraph 24 says "there is no requirement to have a standalone AI policy or governance framework", though documenting how AI is used can help a firm show it meets its existing obligations. Paragraph 25 says regulated entities "may also wish to publish an AI Statement or Policy explaining how AI is used within the organisation".
Paragraph 26 says they "should consider whether their privacy policy or privacy notice requires updating to reflect any use of AI involving personal data". For a starting point on written policy, see our one-page AI policy template.
Margo Legal is not a law firm and does not give legal advice.
Frequently asked questions
What is CILEx Regulation's AI guidance and who does it cover?
CILEx Regulation published its Guidance on the Use of Artificial Intelligence on 1 October 2026. In its own words, it "explains how the existing requirements of the CILEx Code of Conduct apply when AI is used". It "applies to all individuals and entities regulated by CILEx Regulation" (paragraph 4).
Who is responsible for AI-assisted work?
Regulated individuals and entities remain accountable, whether or not AI was used (paragraph 8). Paragraph 10 says "You cannot delegate responsibility for your work to the AI tool or its provider." Professionals "should be able to explain the basis on which AI-assisted work has been reviewed, amended and approved before it is used".
Which uses does it treat as higher risk?
Paragraph 6 separates lower-risk tasks, such as administrative support and summarising documents, from higher-risk uses. Those "may include generating substantive legal analysis, advice, pleadings, witness evidence or material intended for courts, tribunals or regulators", and the guidance says they "require closer scrutiny, verification and oversight".
Does it say clients should be told about AI use?
Paragraph 19 says regulated individuals and entities "should consider whether clients ought to be informed". Disclosure "is likely to be appropriate" where AI "materially contributes to the service provided", where confidential or personal information goes into an AI system, or where a client might otherwise be misled about the extent of human involvement.
Does a firm need an AI policy or an AI Statement?
Paragraph 24 says "there is no requirement to have a standalone AI policy or governance framework", though documenting how AI is used can help a firm show it meets its existing obligations. Paragraph 25 says regulated entities "may also wish to publish an AI Statement or Policy explaining how AI is used within the organisation".
Sources
- CILEx Regulation, "AI use in Legal Services: new CRL Guidance", news, 1 October 2026: https://cilexregulation.org.uk/2026/10/01/new-crl-guidance-ai-use-in-legal-services/
- CILEx Regulation, Guidance on the Use of Artificial Intelligence (footer dated September 2026): https://cilexregulation.org.uk/wp-content/uploads/2026/09/CRL-Guidance-on-the-use-of-Artificial-Intelligence-300926.pdf