Australian law firms and AI answers: what to publish when superlatives are restricted

Why superlatives are a problem for law firms

Australian solicitors cannot advertise in ways that mislead, and "best" or "leading" claims are hard to substantiate. Rule 36 of the Australian Solicitors' Conduct Rules requires that a firm's advertising is not false, misleading or deceptive, or likely to mislead, and is not offensive or prohibited by law.¹ The Australian Consumer Law applies on top.

The rule also covers expertise. A solicitor must not give a false or misleading impression of specialist expertise, and may use the words "accredited specialist" only with accreditation from the relevant professional association.¹ Several states add further limits on advertising personal injury services, so check your own jurisdiction before you publish. Your law society's ethics service can answer specific questions.

So the cautious firm writes modestly. Meanwhile AI answers do not write modestly. Ask an engine for "the best family lawyer near me in Australia" and it will name firms, rank them and describe them in words no firm could publish about itself.

Where do AI answers get their law-firm descriptions?

From whatever they can retrieve: legal directories, review sites, law society listings, news coverage and the firms' own websites. The engine then summarises. When the firm's own site is thin, the summary leans on third parties, and the superlatives come from them.

That has two consequences. First, a firm cannot instruct an AI engine to stop calling a rival "top-rated". Second, the firm's own pages are the one source it fully controls, and they are often the least specific source on the list. A practice-area page that says "we provide expert advice tailored to your needs" gives an engine nothing to quote. A directory entry with a ranking and a review count gives it plenty.

Clients are already asking. Clio's 2025 Legal Trends Report, reported by Canadian Lawyer, covers legal clients who have asked AI legal questions.² Our earlier explainer on US surveys sets out what that research measures and what it does not, including the gap between asking AI and hiring through it.

Australian practitioners are adopting AI on their side too. LexisNexis's latest Australian Legal AI Survey, published in February 2026, covers adoption, confidence and the profession's concern with accuracy grounded in authoritative sources.³

What can a firm publish that AI can cite?

Facts that are true, checkable and specific to your firm. Our view: the compliant page and the citable page are the same page, because both rules reward the same thing. Here is what to put on each practice-area page.

Practice-area pages: what to replace

Hard to substantiate, nothing to quote

• "Leading" or "top" firm claims
• "Specialists in" without accreditation
• "Tailored advice for every client"
• Results presented as typical

Verifiable, citable facts

• Matters you act in, and matters you refer out
• Accreditation held, by whom and in what area
• Fee basis: fixed, hourly or staged, and when estimates are given
• The lawyer responsible, with admission year and jurisdiction

Scope, stated plainly

Say which matters you act in and which you do not. "We act for employers in unfair dismissal claims before the Fair Work Commission; we do not act for employees" is a fact an engine can repeat accurately. It also stops the engine recommending you for work you decline.

Credentials that exist on a register

Name accredited specialists with the area of accreditation, as the accrediting body records it. Give each lawyer's admission year and jurisdiction. These facts can be checked against public registers, which is exactly what makes them safe to publish and useful to an engine matching sources.

How fees work

Explain the fee basis for common matters: fixed fees where you offer them, hourly rates or ranges where you are comfortable, and when the client receives a costs estimate. Fee questions are among the most common questions people put to AI about professional services. A page that answers them, within your disclosure obligations, gives the engine an accurate source in place of a guess.

Process and timing

Describe what happens after first contact: the initial consultation, conflict checks, the engagement letter and typical stages. Avoid promising outcomes. Process facts carry no outcome claim, so they sit comfortably inside rule 36 while answering what prospective clients actually ask.

What about corporate clients?

Commercial firms face a second question: whether clients see them as using AI well. In Thomson Reuters' Future of Professionals 2026 survey of 1,816 professionals in 62 countries (March–April 2026), 78% of corporate clients said AI-enabled quality from their providers was very important or essential.⁴ Only 6% said most or all of their providers delivered it.⁴

That gap is a publishing opportunity inside the rules. A firm can describe how it uses AI, which tasks stay with lawyers, how outputs are checked and how client data is protected. Those are factual statements about process. Our view: a clear, dated page on AI use will answer a general counsel's question better than any claim to be "innovative".

Corporate clients and AI

Clients expect it; few see it delivered

78%of corporate clients call AI-enabled quality from providers very important or essentialThomson Reuters, Mar–Apr 2026 ⁴

6%say most or all of their providers deliver itThomson Reuters, Mar–Apr 2026 ⁴

When AI gets your firm wrong

It will, sometimes: an old address, a lawyer who left, a practice area you dropped, or a superlative you would never publish. You cannot edit the answer, but you can fix the sources. Correct directory and law society listings first, then your own pages, then ask for corrections on review sites where the facts are wrong.

Our guide to fixing what ChatGPT gets wrong about your company sets out that order. For a firm, the main addition is compliance: if an AI answer attributes a claim to you that rule 36 would not allow, record it with the date and engine. You did not publish it, and a dated record shows that.

Then measure. Run the same set of questions, such as "family lawyer in Queensland who offers fixed fees" or "employment lawyer for small businesses in Victoria", across the engines your clients use. Run each on more than one day, because answers vary. Our AI visibility audit runs each prompt on three separate days across six engines.