Litmus publishes its first report: surveying is adopting AI faster than it can explain it

Litmus has published its first report. Faster than we can explain: what surveying is actually doing with AI is the first edition from an independent group set up to look across the whole of the UK surveying at what AI is actually doing to the profession. It is free to read in full, and it can be shared and quoted with attribution.

I should say straight away that I am one of its six contributors, taking part in a personal capacity. So read this as an insider's account of why the group exists and what it heard, rather than a neutral review. Nina Young, founder and chief executive of Surveyors UK, convened the group and chairs it, and she set out some of the findings in Property Week this week.

For surveyors, quantity surveyors and project managers, the report is a useful reality check on how AI is being adopted across the profession. It puts practitioners from residential valuation, cost and standards, consultancy and professional liability in one room and publishes what they see without a vendor or a professional body influencing the conclusions.

In her foreword Nina describes AI arriving across surveying and the built environment "accompanied by a great deal of noise": vendors with something to sell, confident opinions that travelled further than the evidence behind them, and very little that had been tested. Then the RICS professional standard on the responsible use of AI in surveying practice took effect on 9 March 2026, and every firm had to work out what it meant for them.

There is good work happening inside individual disciplines. What was missing, in Nina's words, was "anywhere that looked across the whole profession at once". Litmus is meant to fill that gap. The group meets quarterly under the Chatham House Rule. Contributors are unpaid, take part in a personal capacity, and our firms have no standing in the group. Nothing is published until every contributor has approved it, and no funder, vendor or professional body has any influence over what we discuss, find or publish.

The first session, on 21 July, brought five contributors and the chair together from deliberately different corners of the profession.

Chris is also our podcast guest this week, offering a chance to hear the practical perspective in more detail.

The report groups the discussion into five findings. The first is that the RICS standard produced a scramble, then a lull. Firms completed the visible compliance work in the weeks after March and then stopped, and few have a view of what a second stage involves. I still think the standard has done something important. As I put it in the session, "it has been really useful to bring it to the top table and get people taking this seriously."

It is also being read beyond RICS.

A UK standard becoming a reference point elsewhere raises the stakes for how well it is implemented here.

The second finding is that adoption does not track firm size. Contributors described sole practitioners experimenting seriously and international consultancies avoiding AI altogether, with every position in between. The third is that exposure builds whether or not a firm has decided anything, and Suzanne put that more plainly than anyone.

Staff, subcontractors and clients are already using these tools, so a firm with no position has still taken on the risk.

The fourth finding is the one Nina led with in Property Week: AI is being used against surveyors as well as by them. Nik separated the two: "The first are surveyors who are users of AI. The second is where AI is being weaponised against the surveyor by complainants and claimants." Complaints now arrive well structured, citing specific RICS standards and setting response deadlines. One client submitted AI-fabricated evidence convincing enough to force a re-inspection, and in an internal test described in the report, reviewers could not reliably tell whether a structural defect added to an ordinary interior photo was real.

The fifth finding had nothing to do with technology. The instinct that catches an error before it becomes a claim is built by doing the work, and the group's view is that it is being spent faster than it is being built.

The report is candid about its limits, and I think that is its strength. It is a reading of one conversation, not a dataset, and it makes no claim to represent the profession as a whole. It carries no survey figures because the survey has not run yet. The report makes no attempt to fill that gap. Litmus exists because too many claims about AI have travelled further than the evidence behind them, so its first report applies the same scrutiny to itself.

That survey is the next step. The State of AI in Surveying survey opens in Autumn 2026 and will run across every surveying discipline and level of seniority, from people still qualifying to firm owners. It will be anonymous by default, with no funder influence over the questions or findings, and you can answer for yourself or on behalf of your firm. The report is especially keen to hear from anyone who thinks they have nothing useful to add, because that view is part of what needs measuring. After that, Litmus will publish a report after each quarterly session.

Read the report, then take part in the survey when it opens. Each section of the report closes with a short list of what it means for firms, and together they make a sensible starting checklist for any practice.

The full report is free. It runs to six short sections, five of them closing with what the findings mean for firms, and it is easy to read in one sitting. Download Faster than we can explain from Surveyors UK, share it with your team, and then tell Litmus what your firm is seeing when the survey opens.

All content reflects our personal views and is not intended as professional advice or to represent any organisation.