One in three clients want zero AI in their legal communications. Marketers need to listen

By Moneypenny

The 38% problem: what new AI trust data means for legal marketing

Most conversations about AI in legal services focus on what the technology can do. Fewer focus on what clients want from it. New consumer research gives legal marketers a number worth sitting with: 38% of UK consumers say they want no AI involvement at all in their communications with a law firm.

That is the highest resistance figure recorded across every sector tested, including healthcare, automotive and general customer service. If you are responsible for how your firm talks about AI, whether that is on your website, in pitches, or in new business conversations, this is not a side note. It is the headline.

Where the data comes from

The figures in this piece come from research commissioned by Moneypenny and conducted by Censuswide among 2,000 nationally-representative UK consumers between 8-10 June 2026. Respondents were asked which parts of legal communications, if any, they would be comfortable handling through an AI system.

38% say no parts at all

Across the full sample, 38% selected 'no parts' when asked which elements of a legal enquiry they would be comfortable putting through AI. That compares with lower resistance figures in sectors like automotive and general health and personal care, where consumers were more willing to let AI handle at least something.

For legal marketers, this is the number that should reframe how AI gets talked about externally. It is not a niche concern held by a cautious minority. It is more than a third of the public and, in some demographics, considerably more.

Resistance is not evenly spread

The headline figure hides a wide generational split. Resistance climbs sharply with age:

Generation X sits at 44% and the gap widens further moving into Baby Boomer and Silent Generation respondents, where resistance reaches 50.7% and 58.3% respectively. Younger consumers are markedly more open: roughly one in four 18- to 34-year-olds said no parts at all, less than half the rate seen in older groups.

That gap matters commercially. Client bases skew differently by practice area. A firm with a heavy private client, conveyancing or wills and probate book is likely to be talking to an older, more AI-resistant audience than a firm focused on commercial, technology or start-up clients.

Where clients are more open

The same dataset shows where comfort levels are highest, and it is worth pairing with the resistance figure rather than reading in isolation:

Even the highest-comfort task - an initial enquiry - sits below 30%.

None of the tasks tested come close to majority comfort. The pattern across the dataset is consistent: clients are most willing to let AI handle low-stakes, transactional, front-door tasks, and least willing to hand over anything that touches case substance, advice, or money.

Why this matters for legal marketing specifically

Legal services sit in a different category to most consumer-facing sectors. Clients are often dealing with a stressful, high-stakes or emotionally-loaded situation: a dispute, a property purchase, a family matter, an injury claim. Trust is the product as much as the legal advice itself. That context explains why legal resistance to AI outpaces every other sector measured.

This creates a genuine positioning challenge. Many firms are investing in AI-enabled tools, from intake automation to AI-assisted call handling, because the efficiency case is real. But if marketing communicates that investment the same way a retail or hospitality brand might (leading with automation, speed and innovation), it risks alienating more than a third of the audience before any commercial benefit lands.

The data suggests three things legal marketers should take from this:

• Lead with outcomes for the client, not the technology. Faster response and consistent service matter more to clients than the fact that AI is involved in delivering them.
• Be explicit about where the human is. Given how strongly clients want people involved in anything substantive, naming the point at which a real person takes over is a trust signal, not a weakness to hide.
• Segment messaging by practice area and client demographic. A firm serving an older, more risk-averse client base needs noticeably more reassurance-led language than one serving younger, more digitally comfortable clients.

The framing trap to avoid

There is a temptation, especially in competitive pitches, to talk about AI adoption as a marker of innovation and efficiency in its own right. The data argues against that as a standalone message for client-facing content. Innovation language works well internally and in operational contexts. Externally, with this level of resistance, it can read as a firm prioritising its own efficiency over the client experience, particularly with older or more traditional client segments.

A more durable approach is to position AI as infrastructure that supports the human relationship rather than a replacement for it: AI handles the routine and is predictable so that the people clients actually want to speak to have more time and better information when it matters.

“The data backs up what we are seeing in client conversations every day. People are not asking for more automation, they are asking to feel confident their matter is in good hands. The firms that will win on AI are the ones that use it to protect that confidence, not the ones that use it to look impressive on a pitch deck.”

Bernadette Bennett – Head of Legal at Moneypenny

What this means in practice

For legal marketers reviewing website copy, pitch decks or new business materials, this research gives a useful filter. Before publishing anything about AI in client communications, it is worth asking whether the message would reassure the 38% who want no AI involvement at all, or whether it only speaks to the minority who are already comfortable.

Firms that get this balance right, being clear about where automation adds genuine convenience while being equally clear about where a person remains in control, are likely to build more trust with a wider range of clients than firms that lead with the technology itself.

The 38% figure is a useful corrective for legal marketing teams moving quickly to capitalise on AI 

It does not mean AI has no place in legal client communications. The data on task-level comfort shows there is real, if cautious, appetite for AI handling initial enquiries and routine admin. But it does mean the framing matters enormously. Legal clients are not buying speed for its own sake. They are buying confidence that their matter is being handled properly, by people who understand what is at stake. Marketing that keeps that front and centre, with AI positioned as support rather than substitute, is the version most likely to land.