The Brief
Work & Economy 4 min read

55% of Lawyers Use AI. Only 26% Charge Less for It.

NAVION

Share

More than half of Victoria’s lawyers now use artificial intelligence in their daily work. That figure, drawn from a 2026 census of 2,221 legal practitioners conducted by the Victorian Legal Services Board and Commissioner’s legal services research centre, marks a sharp rise from 37% the previous year. The headline number is striking. What sits beneath it is more instructive.

Adoption Is Broad. The Benefits Are Uneven.

The spread of AI across the legal profession is no longer a story about early adopters. In 2025, usage was concentrated among lawyers under 30. By 2026, that age gap had largely closed, with the notable exception of practitioners aged 60 and over. The technology has moved from a generational preference to something closer to a professional norm.

But adoption does not map evenly across practice areas. Lawyers working in trade practices report AI use at 85%, with banking and finance and information technology and telecommunications both at 78%. Criminal law sits at 29%, family law at 45%, administrative law at 46%. The gap reflects something real: fields that handle high volumes of structured, document-heavy work tend to extract more from AI tools than those requiring nuanced human judgment in emotionally complex situations.

One detail worth noting: 79% of lawyers using AI rely on general-purpose consumer chatbots such as ChatGPT, Claude, Gemini or Copilot. Only 22% use legal-specific tools exclusively. This means the majority of AI-assisted legal work is happening through platforms built for general audiences, not for the specific demands of legal reasoning, citation standards, or jurisdictional accuracy.

Efficiency Gains Are Real. Fee Reductions Are Not.

Here is what most coverage of AI adoption misses. Efficiency and economic benefit are not the same thing, and the gap between them matters enormously for clients.

Only 26% of lawyers reported that their use of AI had led to lower fees for clients. Among those working with business clients, that figure rises to 34%. For lawyers serving individual fee-paying clients, it drops to 23%. The principal researcher and report co-author, Dr Georgie Rychner, offers a clear explanation: corporate clients tend to be repeat users of legal services, they often use AI in their own operations, and they have both the knowledge and the leverage to negotiate lower prices. Individual clients, particularly those in the community legal sector, lack that same capacity.

Rychner also notes that efficiency gains have not been as high as expected overall. One reason may be the verification burden. In 2025, 89% of practitioners reported taking active steps to check the accuracy of AI output. Checking AI work takes time. That time offsets the time saved.

A separate dynamic is adding pressure from the other direction. Clients are increasingly arriving at legal consultations having already consulted AI tools themselves. In 2026, 46% of lawyers reported that clients were bringing inaccurate or misleading information to appointments, and 41% said clients were presenting AI-generated documents that required correction. Nearly half of lawyers said that when clients had used AI beforehand, their own workload increased. Only 19% said it led to a reduction in effort. The technology meant to save time is, in a significant share of cases, creating new work.

What This Tells Us About AI in High-Stakes Professions

The Victorian data points to a pattern that is likely to appear across other knowledge-intensive fields as AI adoption accelerates. The technology spreads faster than the infrastructure needed to govern it, and the efficiency gains accrue unevenly, often to those already best positioned to capture them.

The legal profession is also confronting a specific accountability problem. Courts and regulatory bodies across Australia have been responding to AI-related errors since 2023. A Victorian lawyer became the first in the country to face sanctions for submitting false citations, losing the ability to practise as a principal lawyer. Regulatory investigations followed in Western Australia and New South Wales. The federal court issued new rules on generative AI use. The chief justice of the high court, Stephen Gageler, described AI-generated content in legal cases as having reached an “unsustainable phase.”

These are not isolated incidents. They reflect a structural tension: AI tools are being adopted at scale in a profession where accuracy is not a preference but a legal and ethical obligation. The cost of an error in a medical record or a financial model is serious. The cost of an error in a court submission can affect someone’s liberty, their family, their livelihood.

The 55% adoption figure tells one story. The 26% fee reduction figure tells another. And the 46% of lawyers dealing with clients who arrive misinformed tells a third. Together, they describe a profession in genuine transition, one where the technology is present but the benefits remain unevenly distributed, and where the risks are still being mapped in real time.

In Short

AI adoption among Victorian lawyers has grown rapidly, but efficiency gains have not translated into lower costs for most clients. General-purpose chatbots dominate over legal-specific tools. A growing share of lawyers report that clients using AI independently are arriving with inaccurate information, adding to workloads rather than reducing them. The profession is navigating real accountability risks, with sanctions and new court rules already in place. Adoption is the easy part. Capturing the benefits, and managing the errors, is where the harder work begins.

Based on reporting from The Guardian.

Written by

NAVION