New study finds legal AI has shifted from novelty to core infrastructure, but law firms are failing to discuss efficiency gains with clients
A new study on AI adoption across the legal market, analysed by Law.com's Legal Technology News, finds that legal AI has moved from an experimental tool to the basic operational infrastructure of legal practice. The study's headline finding — characterised in commentary as AI transitioning from "shiny novelty to the basic machinery of legal practice" — reflects a maturation point where questions are no longer about whether to deploy AI but how to govern and maximise its use. The study identifies a significant disconnect: while AI is being embedded in workflows across both law firms and in-house legal teams, law firms — the part of the legal market with the longest-standing client relationships — are the segment least likely to be talking to their clients about the efficiency gains their AI tools are generating. This creates a commercial risk: if firms do not proactively communicate the productivity improvements AI delivers, clients will begin to question whether AI-driven efficiency is being passed through in pricing or retained as margin. A separate analysis from Diginomica examining AI's role in the English legal profession notes that the human professional remains indispensable as a quality-control gate — summarised as the need to put "the human touch not behind the machine, but in front of it, at the gate." The piece references the recent example of Pinsent Masons being criticised for submitting AI-generated references containing hallucinations to a court, then compounding the error by using AI to draft the explanatory letter, which also contained errors. The incident underlines that AI output in contentious and court-facing work requires active professional oversight, not passive review.
Why this matters
The finding that law firms are not discussing AI efficiency gains with clients is commercially material: as AI-driven productivity improvements become visible on the bill (through lower hours for equivalent output), clients will expect either reduced fees or enhanced service scope. Firms that get ahead of this conversation — proactively reframing AI efficiency as a quality and responsiveness benefit rather than a cost reduction — are better positioned to retain pricing power. The Pinsent Masons hallucination incident, cited in the Diginomica analysis, also illustrates a concrete professional responsibility risk: AI-generated court submissions require the same verification standard as manually drafted ones, and failures at that gate carry reputational and regulatory consequences under the SRA (Solicitors Regulation Authority)'s existing conduct framework.
On the Ground
On an AI governance mandate, a trainee would assist with drafting an AI governance policy for a law firm or in-house legal team — setting out permitted uses, quality-control requirements, and escalation procedures for AI-generated output. They would also complete vendor due diligence questionnaires assessing the data security, accuracy, and audit trail capabilities of AI legal tools being considered for adoption.
Interview prep
Question you might get
“If you were a partner at a firm that had deployed AI tools that materially reduced time spent on document review, how would you approach the conversation with a longstanding client about how this affects their billing?”
Sign up free to see the full answer
A model answer you can lift into an interview — how to frame this story for a partner.
Sign up freeSources
My notes
saved