Source: JD Supra

On 22 April 2026 the City of London Law Society hosted a panel exploring legal professional privilege in the age of AI, with Sir Colin Birss delivering the keynote and contributions from Colin Passmore KC, Christina Blacklaws, Jonathan Peddie and Manesh Tanna. The discussion acknowledged recent cautious judicial and professional guidance (including updated Judicial AI Guidance from October 2025 and Bar Council guidance) and described how the judiciary is already using secure AI systems such as a protected Microsoft Copilot instance and an in‑house tool developed with HMCTS and the Ministry of Justice AI Unit. Practical judicial uses cited included transcription, anonymisation of judgments, checking draft judgments for internal inconsistency, finding material in emails and identifying “jigsaw” identification risks where seemingly anonymised details might reveal identities.

The panel turned to legal advice privilege, noting that AI can produce answers that a user would perceive as legal advice—illustrated by a mundane example of a chatbot advising on suing a builder for a broken window—yet such AI‑generated exchanges generally lack the protection afforded to lawyer‑client communications under current English law. Sir Colin referenced the Prudential decision denying extension of privilege to non‑lawyer advisers and declined to endorse changing the privilege test; confidentiality remains a prerequisite for privilege and putting material into public AI systems risks losing that protection. Panelists recommended that lawyers avoid inputting confidential or privileged material into public AI systems, prefer secure controlled systems, expect future disclosure disputes over prompts and outputs, record AI tool use and supervision in workflows, and treat AI governance as a legal risk matter.