FCA authorised · FRN 724952 0117 325 0027Quote & buy →
Aipedia: Chartered legal executives
Munir shows that AI-drafted grounds with fake authorities can lead to the signing supervisor being referred to their regulator. For CILEX immigration practitioners, supervision of juniors' AI use and keeping client documents out of public tools are now central.
Part of: AI in Chartered legal executives
In short
In Munir [2026] UKUT 81 (IAC), grounds drafted by a junior caseworker contained fake authorities and the COLP who signed them was referred to the SRA. The Upper Tribunal said a supervisor who misses a junior's false citations is likely to be more culpable than one who misses their own, and that uploading client letters or Home Office decisions to a public AI chatbot puts them in the public domain. For CILEX practitioners the parallel duties are outcome 1.2, outcome 5.12 and supervision under outcomes 8.2 to 8.4. PI may respond to client loss, but referral and sanctions are usually uninsured.
Immigration practice involves heavy drafting against tight deadlines: grounds of appeal, judicial review grounds, representations, witness statements and skeleton arguments, often for clients with limited English. Caseworkers use generative AI chatbots to summarise Home Office decisions, draft grounds and find authorities. AI translation and transcription are common too.
In UK and R (Munir) v SSHD [2026] UKUT 81 (IAC), promulgated on 17 November 2025, grounds drafted by a junior caseworker cited authorities that did not exist. The COLP who signed the grounds was referred to the SRA. The tribunal made three points that matter to every supervisor:
CRL has no AI guidance, but its position page says the lawyer remains the decision-maker. Under the CILEx Code of Conduct, outcome 1.2 prohibits knowingly or recklessly allowing the court or tribunal to be misled; outcome 5.12 requires confidentiality; and outcomes 8.2 to 8.4 require delegated tasks to be supervised, done by qualified or authorised people and checked by competent people. Munir maps directly onto 8.2: the supervisor stays accountable. The Divisional Court's judgment in Ayinde at paragraphs 7 to 8 adds that lawyers must check AI research, including work prepared by others.
Many immigration clients give instructions in another language, and AI translation and transcription may be used. A mistranslated date, relationship or event in a witness statement can undermine credibility on appeal, and the client may not be able to spot the error in the English version. AI summaries of country information or medical evidence carry similar risks. Have a qualified person check any translation that becomes evidence, read statements back to the client in their own language, and keep the source material on file so the chain from instructions to evidence can be shown.
Illustrative: a CILEX practitioner supervises two caseworkers on an appeals list. One pastes a refusal letter into a public chatbot and asks for grounds of appeal. The grounds read well and cite four cases; the supervisor checks the arguments but not the citations and signs them. Two cases are invented. The tribunal raises the issue at a hearing, the client's appeal is adjourned and costs are wasted, the supervisor is referred, and the client data has been exposed.
If the client suffers loss, such as a delayed or lost appeal, a negligence claim would usually be considered under the PI policy, subject to its terms; the CILEX £2m minimum applies regardless of wording for CRL-authorised entities. Costs of responding to a regulatory referral may be covered under some policies or extensions; see regulatory and disciplinary costs cover. Fines and sanctions are usually uninsured. A confidentiality breach may also involve cyber or data cover. For the wider immigration PI market see PI for immigration advisers.
Apex arranges PI for CILEX practitioners handling immigration work, and a broker can review how your policy treats AI-assisted grounds and regulatory costs. See the AI for CILEX lawyers hub, AI research for CILEX advocates and CILEX lawyers using AI in SRA firms.
Munir concerned an SRA firm, but its reasoning on checking citations and supervising juniors applies to any lawyer conducting litigation or advocacy, and the CILEx Code contains equivalent duties in outcomes 1.2 and 8.2 to 8.4.
Using a public tool is high risk. The Upper Tribunal in Munir said uploading client letters or Home Office decisions to a public chatbot puts them in the public domain, breaching confidentiality and waiving privilege. Outcome 5.12 requires confidentiality.
Potentially. Munir said a supervisor who misses a junior's false citations is likely to be more culpable than one who misses their own. Outcome 8.2 keeps you accountable for delegated tasks.
Fines and sanctions are usually uninsured. Some policies or extensions contribute to the legal costs of responding to a regulatory investigation, depending on the wording. A client's claim for loss caused by the error would usually be considered separately under the main cover.
Tell us about your firm and the work you do. A broker will review how your current PI treats AI-assisted work, or arrange new cover. No fee for a quote. Or call 0117 325 0027.
Start a PI quote Request a callbackApex Insurance Brokers Limited is authorised and regulated by the Financial Conduct Authority. Registered in England and Wales, company number 07014570. This page is general information about how AI is used in the profession and how insurance may respond. It is not legal, regulatory or insurance advice on your circumstances, and it does not guarantee that cover will be available or on what terms.
