In definitive practice guidance issued in October 2026, the Law Society of England and Wales established core regulatory boundaries governing the use of Artificial Intelligence across legal practice, explicitly declaring that “AI is not a legal entity” capable of holding professional duties.
The doctrine of non-delegable professional accountability
As generative and agentic AI tools become embedded in case analysis, document drafting, and disclosure workflows, the Law Society and the Solicitors Regulation Authority (SRA) have clarified the non-negotiable ethical baseline:
- Absolute Personal Liability: Solicitors remain personally and strictly liable for all work, pleadings, and advice produced with AI assistance. Responsibility cannot be transferred to technology vendors or algorithmic models.
- Not a Substitute for Professional Judgement: AI cannot replace the critical evaluative role of a qualified lawyer. Every submission must undergo rigorous human verification prior to client delivery or court filing.
- Severe Sanctions for Hallucinations: Following multiple tribunal warnings regarding fabricated citations and AI hallucinations, unverified filings will trigger severe disciplinary sanctions under SRA Principles.
Client confidentiality and governance standards
The guidance sets out mandatory compliance imperatives for law firms of all sizes:
- Strict Privilege and Confidentiality: Law firms must ensure client data is never ingested into public models or unencrypted cloud environments without informed, express client consent.
- Supervisory Sign-Off: Senior practitioners must implement transparent internal governance frameworks, ensuring junior fee-earners and paralegals maintain documented supervision over automated drafts.
- Transparent Client Disclosure: Firms are advised to maintain open communication with clients regarding the extent and nature of AI deployment in their matters, reinforcing public trust in legal services.