Artificial intelligence has become mainstream in legal practice, but law firms routinely fail to tell their clients when they use it, a survey published by the Law Society Gazette has found.

The findings highlight a growing gap between technological adoption and transparency. Lawyers have embraced AI tools in daily work, yet client communication about that use lags behind.

What did the survey find?

The survey, reported by lawgazette.co.uk, establishes two central facts:

• AI use in legal work is now commonplace rather than experimental.
• Clients are frequently not informed when their lawyers deploy AI tools on their matters.

That combination — mainstream adoption without disclosure — places the burden on firms to reconcile innovation with professional obligations around client care and informed consent.

Why does disclosure matter to practitioners?

Professional conduct rules across jurisdictions generally require lawyers to act with transparency and to keep clients properly informed. Where a firm uses AI to draft, review or research, the survey suggests many clients remain in the dark.

For practitioners, the practical consequence is straightforward. Firms that use AI without a disclosure policy face reputational and regulatory exposure. Client care letters and engagement terms should state clearly:

• which AI tools the firm uses;
• what tasks those tools perform;
• how the firm supervises and verifies AI output.

Confidentiality is the second pressure point. Any AI tool that processes client data raises questions about where that data goes, who can access it, and whether its use complies with data protection obligations.

What does this mean for the profession?

The survey confirms that the debate has moved on. The question is no longer whether lawyers use AI — they do, at scale — but whether they tell their clients and control the risks.

Firms with clear AI policies, documented disclosure and human oversight of machine output are better placed to meet regulatory expectations. Firms without them now have measurable ground to cover.

The disclosure gap identified in the research signals the next frontier of professional regulation: client-facing transparency about AI, not merely internal adoption of it.