The Law Society of England and Wales has published a new AI practice note telling solicitors that artificial intelligence is not a legal entity and that professional responsibility for legal work cannot be transferred to an AI system. The guidance, published on October 21, applies to all solicitors and types of legal practice in England and Wales.
The note is the Society’s view of good practice rather than legal advice, and the Society is a representative body rather than the profession’s regulator. Even so, it is the clearest statement yet from the body that represents solicitors in England and Wales on how existing professional duties apply when AI tools are used in client work.
The Practice Note: Responsibility Stays With The Solicitor
The practice note covers how AI is being used in legal practice, the risks of different types of AI and ways of accessing them, and how existing regulatory duties apply when AI is used. It also addresses AI governance inside firms, including staff training and supervision, and the effects of AI use on confidentiality, privilege, data protection, insurance and cyber security.
Law Society president Brett Dixon said the guidance “clearly states that professional responsibility cannot be transferred to any AI system”. Solicitors and firms should choose their AI tools carefully and always check that legal work meets professional standards, he said, adding that solicitors do not need to be technology experts but do need to understand the risks.
Akber Datoo, co-chair of the Law Society’s technology committee, said the note is intended to help solicitors translate their professional obligations into practical decisions about AI adoption, bringing the relevant information together in one place.
Disclosure, Oversight And The “No Tool Is Automatically Safe” Rule
The guidance advises solicitors to give clients appropriate information about how AI is being used in their matters. It also suggests solicitors can ask clients to disclose where they themselves have used AI to prepare documents or other material submitted to their legal advisers, and it says any client-generated material should be reviewed and verified regardless of whether AI was involved.
No AI system should be assumed safe or suitable for legal work without assessment, the note warns. Free tools may lack the protections needed for confidential or sensitive information, but paid systems require careful scrutiny too: firms are advised to consider terms of service, how data is stored and shared, where AI providers operate, and whether a tool is suitable for the relevant jurisdiction.
The risk of AI systems generating misleading or inaccurate information is highlighted as a significant professional risk. Submissions, documents and representations produced with AI assistance must be appropriately reviewed by a solicitor, with all references thoroughly checked before work is shared externally. The position sits alongside the wider regulatory landscape the profession already works under, from the EU AI Act’s obligations for AI systems to the copyright fights reshaping how AI models are trained and used.
The Junior Lawyer Warning
The note also warns firms about the longer-term effect of AI adoption on the development of junior lawyers. AI can support solicitors and improve the delivery of legal services, the Society says, but it must not replace the opportunities through which junior lawyers build the core skills, knowledge and professional judgement that legal practice depends on.
Dixon put the point bluntly: “Overreliance on AI systems may result in a catastrophic gap of future legal skills and knowledge.” He added that AI has enormous potential, but that “innovation cannot come at the expense of professional integrity or client protection”.
The warning lands as law firms expand their use of legal AI tools for research, drafting and document review — the same week, for example, that the EU Council agreed its Digital Omnibus position simplifying parts of the digital rulebook that governs how such tools handle personal data.
Why It Matters
The Law Society practice note matters because it fixes accountability at the exact point where agentic AI blurs it. As AI systems move from answering questions to drafting submissions and running multi-step work, the temptation is to treat a tool’s output as the tool’s responsibility. The note rejects that outright: the solicitor signs the work, so the solicitor owns it, however it was produced.
For law firms, the practical consequences are immediate — client disclosure conversations, tool procurement checks and review workflows all become part of ordinary professional compliance rather than optional innovation policy. For legal AI vendors, the note raises the bar on what they must be able to show about data handling, jurisdiction and security before a firm can responsibly buy their product. And for clients, it offers a simple assurance: there is always a named professional answerable for the advice they receive.
Background
The Law Society is the professional body representing solicitors in England and Wales, and its practice notes set out its view of good practice on issues from conveyancing to technology. This AI note follows earlier warnings from the Solicitors Regulation Authority about oversight and confidentiality when lawyers use AI, and it arrives as legal AI adoption accelerates across the profession, with tools for research, drafting and review now embedded in everyday workflows at firms of every size.
Source: The Law Society — Using artificial intelligence in legal practice; Artificial Lawyer
0 Comments