How Are Law Firms Using AI in Marketing? What to Systemise and What to Leave Alone

Almost every article you will find on this subject is a list of five tools. Tool lists are easy to write and they are the reason so many firms have spent two years using AI enthusiastically and have nothing to show for it.

The firms getting something out of this are not using better tools. They are using ordinary tools inside a system that decides what gets automated, what a human still has to do, and how anyone would know whether it worked.

This article is that system. It draws on the campaigns and content we run for law firms in Australia, on the published research where it exists, and on the places where the research contradicts what the industry keeps repeating. Where the evidence is thin or comes from a company selling AI, I have said so rather than dressing it up.

How are law firms actually using AI in marketing?

In three places, and it is worth being precise because most of the hype attaches to the one that matters least. Surveys tell you what share of firms have adopted something. They rarely tell you what it is doing, and the answer is narrower than the headlines suggest: drafting and research, some client communication, and a small amount of analysis.

The first is production cost. Research, drafting, editing, repurposing, translation, image work, first-pass analysis. Work that used to be rationed by how many hours your team had is now rationed by how much of it you can check.

The second is how people find you. Long conversational questions increasingly get answered inside the search results or inside an assistant, and the person never visits the ten blue links that used to feed your website.

The third is decision quality, and this is the one that has barely moved. AI is not better than you at deciding which practice area to invest in or whether an enquiry is worth chasing. Treating it as though it were is where firms lose money.

So the honest framing is this. AI changes the cost of producing marketing and the route by which clients arrive. It does not change what makes a firm worth hiring, and it does not do your thinking.

Why has AI adoption stopped translating into money?

Because adoption was never the hard part, and the industry keeps measuring the easy thing.

Clio's Legal Trends Report found 79% of legal professionals using AI in 2025, which is exactly where it sat in 2024 after jumping from 19% the year before (Clio, 2025, 1,700+ legal professionals, North American sample). The step change happened, then it stopped. What has not happened is the commercial return. In Clio's 2026 report on solo and small firms, only 32% of solos and 31% of small firms reported that AI had lifted revenue, against 59% of enterprise firms.

The reason is not mysterious. 86% of solo firms and 78% of small firms have not changed their pricing since adopting AI. Under hourly billing, an efficiency gain that is not converted into either more matters or different pricing is a revenue cut you have given yourself.

The same pattern shows up outside law. Thomson Reuters surveyed 1,816 professionals across 62 countries and found 91% saying their organisation was falling short of AI's potential value, while 34% were using tools nobody had sanctioned (Thomson Reuters, 2026). The number that matters in that report is this one: where a named AI strategy existed, 66% said AI met or exceeded expectations. Where it did not, 22%. In the UK, LexisNexis found 61% of lawyers using generative AI and only 17% saying it was embedded in strategy and operations.

There is also a gap worth understanding between the surveys. The ABA's technology survey put AI use at 30% of US firms, while Clio put it at 79% of individuals. Both are probably right. One is counting tools the firm bought, the other is counting people using ChatGPT on their own account. The distance between those two numbers is your firm's shadow AI problem, and it is the reason a written policy matters before a tool budget does.

In Australia the picture is more conservative again. The Victorian Legal Services Board's 2025 lawyer census, a regulator-run study rather than a vendor survey, found 36.7% of Victorian lawyers currently using AI in practice (VLSB+C, 2026). If you are reading American commentary and feeling behind, you are probably closer to the middle of your local market than you think.

Which parts of your marketing should you systemise first?

The order matters more than the tooling. This is the sequence we work in, and it is deliberately unglamorous at the start.

Most firms want to start at step four because it is visible. Starting there produces a large amount of content pointed at nothing in particular, which is how you end up with a busy blog and the same number of matters.

Start at step one because everything above it depends on knowing what a signed client currently costs. We have written separately about what a law firm lead should cost, and the first surprise for most firms is not the number, it is that they could not produce it.

The second reason to work in this order is capacity. Do one thing until it runs every week without anyone thinking about it, then take the next one. A firm that tries to systemise six things at once systemises none of them, and that is true whether the work is done by your team, an agency, or an outsourced CMO.

What does an AI content system look like in practice?

Not a person typing prompts into a chat window. That is where everyone starts and it does not survive contact with volume.

A working system has five fixed parts. A research step that gathers real sources and the firm's own material. A brief that encodes what good looks like for this firm, including the voice, the reading level, and the things the firm will not say. A drafting step. A human review step that cannot be skipped. And a publishing step that handles the structural work nobody enjoys, which is internal linking, schema, images and the update of anything the new piece makes stale.

The part people underestimate is the brief. The difference between content that reads like every other firm's and content that sounds like yours is almost entirely in what you put in front of the model, not in which model you use. On our own client work, the substance comes from the practitioner, usually a recorded conversation, and the system handles structure and speed. We have set out that whole approach in our AI content strategy.

The part people overestimate is volume. There is a real difference between forty carefully reviewed pieces that each answer a question a client actually asks, and four hundred pieces generated because the system could.

Is AI-assisted content safe to publish?

Yes, within limits that are narrower than the optimists say and much wider than the doom-mongers say.

Google's published position is that it does not ban AI-generated content, and that the test is quality and originality rather than method. Its guidance says that using generative tools "to generate many pages without adding value for users may violate Google's spam policy on scaled content abuse" (Google Search Central). Read the spam policy itself and the definition is method-agnostic: scaled content abuse is when many pages are generated primarily to manipulate rankings rather than to help users. A firm publishing two thousand templated suburb pages is in scope whether a person or a machine wrote them. A firm publishing twenty reviewed practice-area pages is not.

The more useful evidence is about performance rather than penalties. The Content Marketing Institute's 2026 benchmark study of 1,015 B2B marketers found 87% reporting improved productivity from AI, 80% improved efficiency, and only 39% reporting improved content performance, with 12% saying quality had actually got worse (CMI and MarketingProfs, 2025). Their own summary of it is the best one-line description of the failure mode I have seen: AI helps marketers type faster, not think better.

I could not find a single documented case of a law firm site being penalised by Google specifically for using AI. What does exist are the publishing disasters, CNET and Sports Illustrated, both of which were failures of review and disclosure rather than of the technology. That is the risk you are actually managing.

Are your clients really finding you through ChatGPT?

Some are. Far fewer than the headlines claim, and the number of them arriving as measurable traffic is close to meaningless. Both of those things can be true at once, and the gap between them is the single most misunderstood thing in legal marketing right now.

Start with what is observed rather than claimed. Pew Research logged 68,879 searches by 900 US adults and found that when an AI summary appeared, people clicked a traditional result on 8% of visits, against 15% when no summary appeared. Only 1% clicked a link inside the summary (Pew Research Center, 2025). That is a real and substantial loss of clicks.

But look at where those summaries appear. Pew found AI summaries on 8% of one and two word queries and 53% of queries of ten words or more. Long conversational questions get answered in place. Short, high-intent queries mostly do not.

I checked this against live results for legal searches while writing this piece. On "personal injury lawyer melbourne" and on "car accident lawyer near me" there was no AI overview at all: a local pack, then organic listings, exactly as before. On "do i need a lawyer for a car accident" and "how much does a divorce lawyer cost" there was an overview, and the sources it drew on included Reddit, YouTube and the blog posts of individual small firms.

So the damage is concentrated exactly where consumer law firms built their content acquisition model, which is the informational article that used to pull people in early and hold them until they were ready to call.

Now the traffic question. One legal website agency audited its own law firm client portfolio and found AI referrals averaging 0.47% of sessions, with LinkedIn sending 187% more sessions than every AI answer engine combined. Meanwhile a survey of 1,110 US adults by a legal marketing agency found 41.9% saying they would use ChatGPT to research an attorney. Both of those are vendor figures and should be treated as directional, but the gap between them is the finding. People are using assistants to research and then arriving by another route, usually by searching your name or typing your address directly.

One more correction while we are here. The most repeated statistic in legal marketing this year is that 78% of legal queries trigger an AI overview, the highest of any industry. It comes from a public relations release with no published method, and the vertical tracking that does exist puts healthcare, education and business technology above legal. Do not build a budget on it.

How do you actually get cited by AI assistants?

This is where the honest answer is genuinely different from the SEO playbook most firms are paying for.

Ahrefs correlated AI mentions across ChatGPT, AI Mode and AI Overviews against 75,000 brands. The strongest predictors of being mentioned were YouTube mentions at 0.74, branded web mentions at 0.66, and branded anchors between 0.51 and 0.63. Domain rating came in between 0.27 and 0.33, and backlinks lower again (Ahrefs, 2025). Correlation is not causation and Ahrefs say so themselves, but the size of the gap is hard to ignore, and it matches what shows up in the citations: Reddit threads, YouTube videos, and firms that are discussed rather than merely optimised.

BrightEdge adds a second correction. Tracking AI overview citations over sixteen months, it found only 16.7% of citations came from pages ranking in the top ten, with most of the overlap growth coming from positions 21 to 100. Ranking first is not the entry ticket people assume it is.

Put those together and the practical implications are uncomfortable for the traditional legal SEO retainer.

There is a second reason this matters for smaller firms. In a study of 1,094 buyer categories, 53.7% had no brand that assistants consistently named, and once a category owner emerged it held position in more than 90% of month-on-month comparisons. Most practice areas in most cities are currently unclaimed, and they will not stay that way.

What should you never hand to AI?

The clearest evidence on this comes from outside law, from a field experiment run with 758 consultants at Boston Consulting Group. On tasks inside what the researchers called the jagged frontier, creative ideation work, AI users were 25.1% faster and produced output rated more than 40% higher in quality, with the largest gains going to the weakest performers. On a business problem-solving task that required reconciling quantitative data with interviews, AI users were 19 percentage points less likely to reach the correct answer (Dell'Acqua et al, Harvard Business School, 2023).

Same people, same tool, same week. The difference was the type of task.

For a law firm's marketing, the tasks that sit on the wrong side of that line are the ones where a confident wrong answer is expensive and invisible: deciding which matters you want more of, reading whether an enquiry is genuinely viable, setting positioning, and anything that involves a claim about outcomes.

There is a harder finding still. In a controlled study of sixteen experienced developers working on their own repositories, using AI tools made them 19% slower, and after the fact they believed they had been 20% faster (METR, 2025). Small sample, different domain, and I would not lean on it heavily. But it should make you sceptical of every self-reported time saving in this article and everywhere else, including the widely quoted projection that AI will save professionals twelve hours a week. That figure is what people expect to save, published by a company selling legal AI, not what anyone has measured.

What are the best AI tools for law firm marketing?

This is the question everybody asks and the least useful one, so I will answer it briefly and then explain why it is the wrong question.

Most of the work described in this article is done with general-purpose assistants rather than legal-specific products, and that is where the market has moved. Clio's 2025 data found 40% of legal professionals using legal-specific AI tools, down from 58% the year before, while overall use held steady. People are reaching for the general tools because they are better at the writing and analysis tasks that marketing actually involves.

The categories worth having are small. Something that drafts and researches. Something that handles transcription, so a practitioner's expertise can be captured by talking rather than writing. Something in your website or CRM that scores and routes enquiries. And whatever your ad platforms already give you, which is mostly automated bidding you are using whether you decided to or not.

The reason the tool question misleads is that every firm in your market has access to the same list. Nothing on it is a competitive advantage. The brief you put in front of the model, the practitioner knowledge you feed it, and the review step you refuse to skip are the parts your competitors cannot copy from a blog post.

Where does AI create compliance risk in your marketing?

Almost nowhere that the AI-specific rules cover, and in one place they do not.

Of the fifteen or more US jurisdictions that have issued AI guidance for lawyers, only Florida and California address advertising, marketing or intake at all. Florida's opinion is the clearest statement anywhere: a lawyer must tell prospective clients they are communicating with an AI program rather than a lawyer or employee, and warns specifically about a chatbot that gives legal advice, fails to identify itself, or lacks clear disclaimers (Florida Bar Ethics Opinion 24-1). In the UK, the SRA's 2026 warning notice on misuse of AI is about court submissions, verification and client data, and says nothing about marketing. Australia's regulators have issued a joint statement on confidentiality, verification and billing, again aimed at practice rather than promotion.

So the AI rules are watching the legal work. The exposure in marketing sits where it always did, in the ordinary advertising rules, and AI makes those far easier to breach because it lets you publish everywhere at once.

Australia gives the sharpest example. Advertising for personal injury work is restricted by statute in several states, with Queensland's regime among the tightest, and we have covered what Queensland firms can and cannot say separately. Earlier this year a Queensland tribunal dealt with a firm whose billboard, website and social content used a common form of words. The conduct was upgraded from unsatisfactory professional conduct to professional misconduct, and the penalty was increased from an agreed $2,000 to $30,000 with a public reprimand. One phrase, replicated across channels, over several years.

Now imagine that phrase sitting in a content brief or an ad template that a system uses two hundred times. The firm wears the finding, not the tool.

Two other things belong on the same page. Clio's consumer research found 78% of clients want AI use disclosed, while 35% of lawyers say they rarely or never disclose it, and 36% of consumers said they would be less likely to trust a lawyer who uses AI. And if you are running an intake chatbot, assume you must identify it as one.

How should AI change your intake, and how far should it go?

This is where the money is, and it is also where firms overreach fastest.

The case for automating here is strong because the problem is a timing problem. Enquiries from advertising do not behave like referrals. They arrive cold, at night, from people contacting several firms, and the firm that answers first usually signs them. We have set out the evidence on response time elsewhere, and it is the single most reliable improvement available to most firms.

What AI does well here is everything that surrounds the conversation. Acknowledging an enquiry instantly at 11pm. Capturing the facts that decide viability before a human reads it. Scoring and routing against your written qualification standard. Drafting the follow-up sequence. Transcribing and summarising the first call. Flagging the enquiries nobody has touched in 24 hours.

What it should not do is decide whether a matter is viable, or talk to a prospective client in a way that could be mistaken for advice. The first is a judgment task on the wrong side of the frontier. The second is the Florida problem.

There is also a quieter risk. The firms that get the most out of automated intake are usually the ones with the worst manual intake, and automation can hide that rather than fix it. If nobody currently rings enquiries back three times, an AI system that sends three emails will produce a better-looking dashboard and the same number of signed matters.

How do you measure whether any of this worked?

By the only number that survives contact with a partner meeting, which is what a signed matter costs you and what it is worth.

That means the chain has to be joined up: enquiry, source, qualification decision, signed or not, value. Most firms have the first two and none of the rest, which is why so many agency relationships end in an argument about lead quality. We built PixelRush HQ to close that loop for our own clients, and the honest reason is self-interested as much as anything, because without the feedback from intake we would be guessing like everyone else.

For the AI-specific work, add three measures that most firms do not keep.

The branded search measure is the one I would add first if you are worried about AI search. You cannot see most of the assistant conversations that mention you, but you can see the people who go looking for you by name afterwards.

Will AI replace lawyers, and does it change how you market?

It is the most searched question in legal AI by a distance, your clients are asking it, and the honest answer matters for your marketing rather than for your career.

What the evidence supports is displacement of tasks, not of lawyers. The field experiment with 758 consultants found large gains on generative work and a measurable drop in accuracy on judgment work. The legal-specific version of that distinction is visible in the hallucination case tracker maintained by Damien Charlotin, which had logged more than 2,000 court decisions worldwide by September 2026 where a judge found or implied reliance on fabricated AI content, including over 100 in Australia. Of those, more than 800 involved lawyers rather than self-represented litigants.

So the risk to the profession is not that AI does the work. It is that people who are not lawyers believe it already has.

That is the part that changes your marketing. A share of your prospective clients now arrive having asked an assistant about their problem, and some of them arrive believing they no longer need you. Clio's consumer research found 14% of consumers had used AI to answer a legal question, rising to 26% among Millennials. Firms that treat this as an insult lose those people. Firms that engage with what the assistant told them, correct it where it is wrong and explain what it could not know, convert them.

One of our clients sees this directly. Their intake team reports AI-referred enquiries arriving warmer and better informed than any other channel, and also taking longer to sign, because those people have built themselves a shortlist before they call.

What does a lean team with AI genuinely beat a large team at?

Not everything, and pretending otherwise is how firms get talked into a strategy they cannot sustain.

A national firm with a large marketing department and a broadcast budget will outspend you, out-rank you on the terms everyone wants, and appear in more places than you can. None of that is changed by a subscription.

What a small firm with a working system does beat them at is specificity and speed. A brief that captures one practitioner's actual expertise produces content a national content team cannot match, because the national team is writing for eleven practice areas and cannot get their senior people on a call. A firm that decides something on Tuesday can publish it on Tuesday. And the category ownership research suggests most practice areas in most cities are currently unclaimed in the places assistants look.

The BCG study points the same way for a different reason. The largest gains went to the weakest performers, which means AI compresses the gap between a small team and a large one on production tasks, while leaving the gap on judgment tasks exactly where it was. So the correct strategy for a small firm is to use AI aggressively on production and to spend the time it frees on the judgment work the big firms are too slow to do well.

Where should you start if you are doing none of this?

Pick the smallest thing that gives you a number.

If you cannot currently say what a signed client costs you, start there, because every other decision in this article is unanswerable without it. If you can, start with intake speed, because it is the cheapest gain in most firms. Only then move to content, and when you do, cap it at what your review capacity can genuinely carry.

Then write the two pages nobody writes: a one page AI policy that says which tools your staff may use and what may never be put into them, and a content brief that encodes your voice, your practice areas and the advertising rules you work under. Those two documents are worth more than any tool you will buy this year, and they are the difference between the 66% who say AI meets expectations and the 22% who do not.

The pattern across every credible piece of evidence here is the same. The technology is not the variable. The system around it is, and so is the honesty of the measurement underneath it. That is unglamorous, and it is also why so few firms have done it, which is the opportunity.

If you want to see how we run this for firms rather than read about it, the growth system page sets out the mechanics and our case studies show what it produced.