A survey has revealed nearly two thirds of fee earners admit to using open source artificial intelligence tools such as ChatGPT, potentially compromising client data and confidentiality.
The research by legal software company Access Legal revealed 59% of the 200 UK legal professionals surveyed by Censuswide said they had used unapproved AI applications, like the free version of ChatGPT, for client work, in direct contravention of the SRA Code of Conduct and despite the recent warnings given in a landmark tribunal case.
The Upper Tribunal decision in Munir v Secretary of State for the Home Department [2026] UKUT 81 (IAC) confirmed the use of open AI applications permanently waive legal professional privilege and breaches client confidentiality. The Tribunal did not rule that AI has no place in legal practice, but it drew a clear distinction between open source public tools, and closed enterprise-grade environments where data remains within the firm’s control.
As part of the survey, 100 UK-based legal leaders (managing directors and practice managers aged 30 and over) were asked about the oversight of AI within their firms: 68% said they have full visibility of AI use at their firm and believed they face ‘zero risk’ of unapproved AI use. The Tribunal was explicit that conduct of this kind warrants referral to the SRA and must in any event be reported to the Information Commissioner’s Office.
The research exposes a disconnect that has serious regulatory consequences, Access Legal warns. Under the SRA Code of Conduct, firm leaders have a direct duty to supervise. A leader who believes their firm has zero exposure while fee earners are actively using the tools the Tribunal has now ruled waive privilege is in breach of that duty.
Almost three quarters of the paralegals surveyed (71%) and 57% of solicitors admit to using unapproved AI to keep up with workload demands, which Access Legal says illustrates the discrepancy between the tools fee earners want and those that are being provided by firms. Half of fee earners surveyed said they want AI built into their case management system, yet only 25% of respondent firms currently have that feature in place – a gap Access Legal says is driving the use of unapproved alternatives.
Andrew Stevens, general manager at Access Legal, said: “What has surprised us is the sheer volume of legal professionals who are using unapproved AI tools and the fact the research strongly indicates a significant portion of leaders are unaware. Everyone understands that practitioners look for faster ways to work; but following Munir, the consequences of reaching for the wrong tool are no longer theoretical, they are regulatory, permanent, and well-documented.”
For firm leaders, the question is no longer whether shadow AI use is happening, but whether leaders know about it and what, if anything, they are doing about it, Access Legal suggests.
Stevens added: “Firms that want to get ahead of this need to act on two fronts. The first is technology. Giving their people approved, secure tools that meet the demands of the job removes the incentive to go elsewhere.
“The second is culture and supervision. Having open conversations at every level about what people are using and why, and putting clear oversight in place. Firms that do both are in a far stronger position. Firms that do neither are exposed in a way that, post-Munir, they can no longer afford to be.”
