Adoption is broad, but uneven
Use varies sharply by practice area. Lawyers in commercial law lead at 85%, followed by banking and finance, and information technology and telecommunications, both at 78%. Use is lower in criminal law at 29%, family law at 45% and administrative law at 46%.
Most are turning to general-purpose chatbots rather than legal-specific products:
The age gap in AI use has almost disappeared, though adoption remains low among lawyers aged 60 and over. In the 2025 survey, lawyers under 30 were more likely to use the technology.
Lower fees are not the main return
Fee reductions remain the exception. Among lawyers working with companies, 34% reported lowering fees after adopting AI; among those serving private clients for a fee, the figure was 23%. Another 18% said AI had made it viable to take on cases that would previously have been uneconomic.
Chief researcher and report co-author Dr Georgie Richner said the greater pressure to reduce fees in corporate work and large firms may reflect clients’ regular use of legal services and familiarity with them. Some clients also use AI themselves, making them better able to assess its effect on costs and push for lower prices. Researchers saw no comparable effect in community legal assistance.
The efficiency gains, Richner said, were smaller than expected. In the 2025 survey, 89% of practising lawyers said they checked AI’s answers for accuracy—a step that can eat into the time saved.
Clients bring the cleanup work
In 2026, more clients were using AI to conduct their own legal research. Some 46% of lawyers said clients arrived with inaccurate or misleading information, while 41% said they received AI-generated documents that needed correction.
Nearly half said working with clients who use AI increased the time and effort involved; 19% said it reduced them. Richner said lawyers more often encountered mistaken claims about clients’ legal problems than evidence of deeper understanding. Assessing clients’ own views, she added, will require further research.
Courts are setting limits
Since 2023, Australian courts and legal professional bodies have been responding to rising AI use by lawyers and the public. Last year, a Victorian lawyer became the first in Australia to be disciplined over false case citations discovered in legal work; the lawyer was barred from working as a principal lawyer. Regulators opened investigations into similar cases in Western Australia and New South Wales.
In April, the federal court issued new rules for using generative AI. Last year, High Court Chief Justice Stephen Gageler said the use of AI-generated material in court cases had reached an “unsustainable point.”
I think the central tension is not whether AI is entering legal practice; the numbers suggest it already has. It is whether the time spent checking outputs and correcting clients’ AI-assisted work will leave much efficiency to share—and whether lower fees will follow even when clients know enough to demand them.
Daily AI news
Every day we pick what actually matters in AI and explain it plainly — no hype, no filler. Subscribe if you want to follow where the industry is going.
Only what matters — every day
Follow on X