The Law Society of England and Wales has published a practice note on the use of artificial intelligence in legal practice, setting out how solicitors’ existing professional duties apply when AI tools are used. The guidance was issued on 21 October 2026, according to trade reports, and is available on the Law Society’s website.
What the practice note says on confidentiality and tools
Legal Cheek reported on 22 October that the note warns free AI services often lack adequate protections for client confidentiality and data security, and that even paid-for tools should be checked before use. Today’s Wills and Probate, which quoted the guidance directly, reported that free tools “often do not offer the necessary protections”.
The same report said firms are told to consider how they access AI systems, the terms of service, how data is stored and shared, and where the AI provider operates. Legal Cheek said the note also states that no AI tool should automatically be considered safe.
Accountability and supervision
On responsibility, the Law Society president Brett Dixon was quoted by Today’s Wills and Probate and by Artificial Lawyer as saying that “professional responsibility cannot be transferred to any AI system”.
City AM reported that the note says professional duties “apply in the same way whether AI or other technologies were used” by the solicitor or “by anyone under that solicitor’s supervision”. Legal Cheek said solicitors remain accountable for work delivered to clients, courts and third parties, and that AI-generated documents, submissions and references need human oversight and checking, particularly for inaccurate information and fabricated case citations.
The note also covers transparency. According to Legal Cheek, solicitors should tell clients how AI is used in their matters, and firms can ask clients to disclose when they have used AI to prepare materials.
Mr Dixon was also quoted by Legal Cheek as saying: “Solicitors do not need to be technology experts, but they do need to understand the risk.” He added, in the same report: “AI has enormous potential, but innovation cannot come at the expense of professional integrity or client protection.”
Training and junior lawyers
Several outlets reported that the guidance addresses the development of junior lawyers. City AM said the note calls on firms to make sure AI does not replace training opportunities, including learning to “identify errors, omissions and unsupported propositions”. Mr Dixon was quoted as saying: “Over-reliance on AI systems may result in a catastrophic gap of future legal skills and knowledge.”
Akber Datoo, chief executive of D2 Legal Technology and co-chair of the Law Society’s technology and law committee, was quoted by Today’s Wills and Probate as saying: “Good governance should give firms the confidence to move forward, not an excuse to stand still.”
Artificial Lawyer described the Law Society’s guidance as voluntary and said it carries no legal authority. The Law Society is a representative body rather than the regulator: solicitors are regulated by the Solicitors Regulation Authority.
Research on unauthorised AI use
Separately, AI Score, a company that works with organisations on visibility and control of AI use, has published research on how employees are using AI at work. It said almost a quarter (24%) of UK organisations have identified employees using unauthorised AI tools.
Among organisations reporting an AI incident or near miss, 44% said confidential or sensitive data had been entered into an AI tool, according to AI Score. The company did not disclose in its material how many organisations took part, which sectors they were in or when the research was carried out, and the figures cover UK organisations generally rather than law firms specifically.
AI Score said the findings suggest many organisations are finding it difficult to keep track of how employees are using AI.
The professional rules behind it
Solicitors’ confidentiality duties come from the SRA Codes of Conduct, which include a rule on confidentiality and disclosure that applies equally when AI is used, according to a Travers Smith article published on the Law Society’s risk and compliance pages in May 2025. The same article said firms adopting AI should keep records of the steps taken to comply with the Codes, and should carry out “regular audits, to assess functionality and effectiveness”.
