Legal technology spent a decade buying efficiency: contract review, discovery, research, compliance. The next line of legal budget goes somewhere the sector has barely priced, which is AI Search visibility for law firms. The uncomfortable part is that this layer is not empty and waiting to be earned. In legal, it is already occupied.

  • Visibility in legal discovery is moving from ranking position to citation inside a generated answer, and cited slots are scarcer than search results ever were.
  • First-party 2026 citation research finds legal directories, not law firms, hold the first cited source in the large majority of lawyer-hiring answers.
  • Publishing more thought leadership is the 2006 SEO playbook relabelled: necessary, non-excludable, and too slow to displace an incumbent citation.
  • The paid lane already exists. OpenAI began placing sponsored units below ChatGPT answers on 9 February 2026.
  • Directories and legal publishers now supply the answer without any structural claim on what the answer earns.

What AI Search visibility for law firms actually means

AI Search visibility for law firms is the share of AI-generated answers in which a firm is named or cited when someone asks an engine for legal guidance or for a lawyer. It is not a ranking, and it is not a session in analytics. It is presence inside the synthesized answer that has replaced the results page for a growing share of queries.

The scale of the substitution is documented. SparkToro's June 2026 analysis of Similarweb panel data found that 68% of US Google searches between January and April 2026 ended without a click. The decision is increasingly made before any firm's website is involved.

The ranking economy and the recommendation economy, side by side

The old legal acquisition economy was a ranking economy with a paid lane bolted alongside it. Ten organic slots, an ad auction above them, and a click as the unit of value. Legal paid more for that unit than any other vertical: WordStream's 2026 Google Ads benchmarks, drawn from more than 13,000 US campaigns run between April 2025 and March 2026, put Attorneys and Legal Services at the highest average CPC of any industry tracked, at $9.87, with high-intent personal injury terms in major metros running into the hundreds.

The recommendation economy has one answer, a handful of grounding sources, and no click to price. Scarcity is the defining feature. Similarweb data reported by TechCrunch on 27 July 2026 found that just 6.8% of US ChatGPT desktop queries returned a cited source as of May 2026. Ten ranked positions have become three to five citations, and legal firms are competing for them against aggregators built to be the canonical list.

What the UnoSearch column argued, and what the citation data shows

Writing in MediaNews4U on 27 July 2026, Pankaj Srivastava, founder of UnoSearch, argues that legal tech's next battleground is visibility rather than efficiency, that AI has inserted a recommendation layer between client and web, and that firms should respond with authoritative content, recognisable expert profiles, and presence in trusted third-party publications and directories. The diagnosis is correct and early. The column carries no data, so we checked the citation layer it describes.

The evidence is not unanimous, and it is worth being explicit about that. InterCore Research, publishing on 20 July 2026, ran 540 lawyer-hiring queries three times each and classified 18,900 citations: a legal directory was the first cited source in 77.8% of answers, and directories took at least 51.8% of all citations. A 5W and Haute Lawyer Network report from 29 April 2026 reached the same conclusion, describing roughly seven ranking sites as a citation cartel across the query categories it tested. A smaller June 2026 study by Taqtics, covering 236 citations from a single grounded engine, inverted the finding and put firms' own sites at 67%. We weight the larger, multi-engine samples, and note that any single-engine snapshot in this market has a short shelf life.

The strongest case for expertise-led content

The case for the earned route is genuinely strong in this vertical. Law firms already produce exactly what retrieval systems reward: structured explanation, regulatory analysis, FAQs, practice commentary. Legal is also the one industry where the paid route carries professional-conduct risk, which pushes rational firms toward authority first. The American Bar Association's own commentary on Model Rule 7.2 warns that a lawyer must not pay a lead generator that implies it is recommending the lawyer without payment or that it has analysed the person's legal problem. Anyone building paid placement into legal answers has to solve labelling before they solve pricing.

Why an incumbent citation does not yield to more content

Authority is a floor because it is non-excludable. Every competitor can publish, every competitor is being told to, and none of it removes Justia or Super Lawyers from the first cited position. In a ten-slot world, a good firm outranked a directory some of the time. In a three-citation world, the aggregator is the structurally correct source for a list-shaped question, and no volume of bylines changes the shape of the question. Meanwhile the same Rule 7.2 commentary that constrains implied recommendations expressly permits paying the reasonable costs of advertising, including online directory listings and internet-based advertisements. Disclosed, labelled paid placement is the compliant path. Undisclosed influence over the answer is the one that ends in a bar complaint.

What this means for law firms and for the publishers that feed the answers

Run two budgets, not one. Keep the authority programme funded, because it is the entry ticket, then ringfence a separate line for placement inside AI answers and measure it on share of cited answers per practice area and metro, not on sessions. A firm paying $150 for a click in a market where two thirds of searches never produce one is already funding the wrong lane. Audit which directories currently hold your practice-area citations, because that is your competitive set now, and it is not the firm down the street.

The advice given to law firms is, read from the supply side, an instruction to extract free rent from you. Directories, legal trade press, and recognition sites are the properties grounding these answers, and their commercial models were priced on visits and profile subscriptions that the answer intercepts. Cited more, paid less, is not a business. The correct response is to price the citation itself: treat your presence in AI answers as sellable inventory, with disclosed sponsored placements and a revenue share, rather than waiting for referral traffic to come back.

Three signals to watch before mid-2027

Three developments will decide how this settles. First, the paid layer is no longer hypothetical: OpenAI started showing sponsored units below ChatGPT answers on 9 February 2026, with Sam Altman stating publicly that the company would not take payment to influence the answer itself, which leaves the space beside the answer as the monetizable surface. Second, watch the first state bar ethics opinion addressing paid placement in AI-generated legal recommendations, because legal will be the vertical where labelling standards get written. Third, watch directory pricing: the moment Chambers, Avvo, or Super Lawyers sells a citation-share package rather than a profile, the category has a price.

Conclusion

Hold on to this: AI Search visibility for law firms is a market with incumbents already in place, and earned authority alone will not evict them. What is missing is the layer that made the old search economy function, a disclosed, priced, measurable ad lane that pays the sources grounding the answer. That is the category Smalk AI operates in: generative engine advertising, placing native ads for AI agents while opening a revenue stream for the media sources whose content those agents cite. Watch the first legal directory to sell citation share as inventory, because that is the moment the answer gets a price and everyone else starts paying it.