What does the SRA Warning Notice require?
The SRA’s Warning Notice identifies two specific areas of concern for every regulated firm.
First, court documents must not contain false information generated by AI. The notice references four cases where AI hallucinations appeared in legal submissions, including R (Ayinde) v Haringey LBC [2025] EWHC 1383 and UK v SSHD [2026] UKUT 81 (IAC). In the Upper Tribunal case, the court reminded lawyers that they remain personally responsible for verifying all content in legal submissions, regardless of the tools used to draft them.
Second, client confidentiality must not be compromised through AI tools. Entering client data into public AI systems without appropriate safeguards risks breaching both professional duties under SRA Principle 6 and data protection requirements. The notice makes clear that existing obligations apply directly to AI usage.
How many AI misuse reports has the SRA received?
The SRA received 42 reports related to potential AI misuse in a single year. That number will grow. Every firm using generative AI in any part of its practice, from drafting to research to document review, now operates under explicit regulatory scrutiny.
The obligation applies to every regulated firm and every individual solicitor. Supervision requirements under Rule 9.4 demand that regulated work is overseen by someone with appropriate experience, and that obligation extends to AI-assisted work.
What are most law firms missing on AI governance?
Most law firms lack three things the SRA now expects.
A documented AI governance framework. Not a policy buried in an intranet folder, but an active framework recording which AI tools are in use and who is accountable. The AI Governance Healthcheck provides a structured starting point for firms that do not yet have one.
Supervision that covers AI outputs. The SRA’s expanded supervision guidance (June 2026) is explicit: if AI-generated content reaches a court, the supervising solicitor is accountable.
Confidentiality controls for AI tools. Ecsper Practice provides the governed workspace where AI operates within defined boundaries, with every action logged and auditable.
What should law firms do now?
SRA Principle 2
You must act with integrity. This includes ensuring that the work you produce, and the advice you provide, is accurate and reliable, regardless of whether AI was used in its preparation.
Start with three steps. First, check your governance exposure in three minutes. Second, document every AI tool in use and assign a named owner. Third, ensure every AI-assisted output has a qualified reviewer before it leaves the firm.
The Warning Notice is not the end of the conversation. It is the beginning of a regulatory framework that will only tighten. The firms that treat this as a moment to build governance infrastructure, rather than a compliance checkbox, will be the ones that use AI with confidence.
See the full tracker: over 2,000 documented cases of AI misuse in court proceedings. Ecsper AI Risk Intelligence
