Client pressure drives law firms' AI investment choices

Litera has published survey findings suggesting that client pressure is driving law firms' AI investment decisions. The research found that 85% of firms are already feeling, or expect to feel, direct pressure from clients over their AI strategy.

The results point to a market in which client expectations are shaping technology spending more directly than internal financial measures. More than half of respondents said a client had directly influenced an AI investment decision in the past 12 months, while only 15% said AI investment remained entirely internally driven.

That shift appears to be pushing firms to focus less on buying access to AI tools and more on showing clients what those tools deliver. Some 32% said they could not confidently demonstrate AI value to their most important client, suggesting a gap between investment and client-facing proof of benefit.

The survey also found that return on investment carries less weight than other factors in legal AI decisions. It ranked last in two separate questions, with respondents placing greater emphasis on time recaptured than on cost avoided when explaining value.

Together, the results suggest the legal AI debate is moving beyond simple procurement questions. Firms are increasingly being judged on whether they can embed AI into day-to-day work, train lawyers and staff to use it, and make a credible case to clients that the technology improves service.

Adoption remains a notable hurdle. Some 36% of respondents said adoption, training, and culture represented the biggest gap in their AI strategy, making internal execution a bigger concern than the underlying tools themselves.

People first

When asked what would distinguish their firm if all competitors had access to the same underlying AI, respondents did not put technology first. Instead, people, talent, and expertise ranked highest at 24%.

Custom workflows and use cases followed at 18.7%, while proprietary data and knowledge came next at 13.3%. The ranking suggests firms see differentiation coming from how they organise work and apply judgement, rather than from access to a single model or software product.

The findings align with a wider pattern in professional services, where clients are pressing advisers to explain how AI changes the quality, speed, and cost of work. In legal practice, that pressure can carry particular weight because client relationships often depend on trust, pricing transparency, and evidence that advisers are using new tools responsibly.

Avaneesh Marwaha, Chief Executive of Litera, linked the findings to broader client doubts about whether AI is producing a tangible improvement in legal service. "I started The Changing Lawyer nearly a decade ago because I believed the profession deserved a space to think about where it was headed, not a place to sell. This State of Legal AI research reflects that. Without necessarily asking it out loud, clients are all wondering: is my lawyer actually better because of AI?" he said.

Client demands

The pressure described in the survey was echoed by Casey Flaherty of Baretz+Brunelle. "Client pressure is no longer theoretical, and firms are right to focus less on model access than on operational execution. In this market, unarticulated value is invisible value," he said.

His comments reflect a central theme in the data: firms may no longer gain much advantage simply by adopting AI, because clients increasingly assume access to such tools is becoming commonplace. The harder task is translating AI use into outcomes that can be articulated in client conversations, pitches, and ongoing service reviews.

That may also explain why respondents placed so much emphasis on human factors. If rivals can buy similar systems, implementation, legal judgement, training, and workflow design become the main areas where firms can still stand apart.

Eric Friedman, former chairman and chief executive of Skadden, Arps, Slate, Meagher & Flom and a Litera board member, framed the issue in commercial terms for firms facing increasingly demanding clients. "Both the State of Legal AI: Spring 2026 Market Sentiment Report data and broader market surveys suggest that clients increasingly expect to share in the AI productivity dividend, and that they have the leverage to ask for it," he said.

The survey was published through the revived The Changing Lawyer publication. Litera, which sells legal technology and AI software, says it serves more than 15,000 customers globally and 2.3 million daily users, including 99% of the Am Law 100.

For many firms, the results suggest the next phase of legal AI will be defined less by access to new systems than by whether lawyers can show clients that the technology is changing legal work in ways that matter. Among respondents, 32% said they could not confidently demonstrate AI value to their most important client.