Artificial intelligence (AI) dominates conversations across the surveying profession, but far less is said about how surveyors are actually using it day to day.
Since the RICS professional standard Responsible use of artificial intelligence in surveying practice came into effect on 9 March 2026, the question for members has shifted.
It is no longer simply what AI can do, but rather how to use it without compromising the standards expected of a surveyor.
The standard is mandatory for all RICS members and regulated firms. As a result, surveyors must understand the limitations and failure modes of the AI they use, and make a written reliability decision on any output that materially affects their service.
To find out what that looks like in practice, the authors spoke to two surveyors who have built AI into their reporting process.
Ashley Humphreys AssocRICS is a residential building surveyor working as a sole practitioner in East Sussex, while Anthony Spencer MRICS is managing director of AMS Surveys, an SME practice based in Liverpool.
Both use AI assistants within the same surveying platform: one to help interpret inspection photographs and another to turn site notes into structured report content.
While Ashley's and Anthony's experiences are very different, their conclusions about using AI are remarkably similar.
The sole practitioner: more time looking at the building
As a sole practitioner, Ashley handles everything himself: the inspection, the client, the evidence and the final report. The most notable change to his working day is in his reporting time. A report that once took approximately five hours now takes between one and a half to two.
That change began on site. Earlier this year, Ashley started dictating detailed notes as he moved around each property, which the AI assistant he uses then turns into draft report sections.
'If you put good-quality data in, together with your knowledge as a surveyor, you get a great output,' he says. 'But the AI still needs the surveyor to put the information in.'
Dictating as he goes has also changed how Ashley inspects: he finds that he looks at the property more closely than before.
'You're seeing more of the building, so there are fewer opportunities for mistakes.'
Another AI tool helps him to interpret his photographs, but he is clear about its limits. Shadows, distance and image quality all affect what can be read from a photo, and the system qualifies its observations to what is visible.
For a sole practitioner, the time saved means more than productivity. Ashley's working day used to be squeezed by the school run at 3.30pm. Now, with draft text in place to check that pressure has eased.
'I feel more comfortable going to the playground. I don't feel under pressure. It's had a really positive impact on my life.'
Figure 1: Ashley Humphreys AssocRICS on site in East Sussex. Source: Ashley Humphreys
The SME: the challenge was people, not technology
From the outset, AMS Surveys has embraced innovation with a simple aim: to keep surveyors surveying and to cut as much admin work as possible.
Anthony began experimenting with general-purpose AI tools around two years ago. He was always conscious of what should and shouldn't go into open models, and it was a platform designed specifically for surveying that persuaded him to adopt AI across the practice.
He expected the hard part to be learning how to use the technology, and whether it could write the way that he and his AMS colleagues do.
'I thought it would be more about prompts and guardrails. It's flipped round. The system was quick to set up. The difficulty has come from bringing the team with us.'
Report writing is personal. Surveyors take pride in how they craft a report and in the wording that protects both client and practice.
Early objections from Anthony's colleagues tended to sound the same: 'The software produces words, but I have to rewrite them myself.'
The surveyors who have embraced AI have changed how they work on site, recording more detail in their notes so the output needs less work in the office. Professional knowledge still fills the gaps.
Anthony gives a simple example: every surveyor knows that DG means double glazing, but the AI won't always. Misread abbreviations and gaps in the notes carry straight into the draft.
He also sees value in AI as a quality check alongside human review, spotting inconsistencies such as a fireplace described in a property where no chimney stack has been mentioned.
It can only flag what is in the notes, however, not what has never been recorded. 'I never want to get away from a second set of eyes on a report,' he says.
AMS is now running a practice-wide pilot to gather feedback. Anthony's longer-term aim is to give surveyors more time with clients, explaining findings by phone or video call rather than leaving them to interpret a dense report alone.
Figure 2: Anthony Spencer MRICS on site. Source: AMS Surveys
Where professional judgement takes over
Both surveyors featured in this article draw the same line: surveyors must assess the reliability of AI outputs and remain accountable for all their work.
Under the RICS standard, the reliability decision must be prepared by, or under the supervision of, an appropriately qualified and named surveyor who accepts responsibility for its use.
The platform both surveyors use keeps a full audit trail every time its AI is used: what went in, what came out and who used it, so it is clear which content was and wasn't AI-generated. Report sign-off is recorded in the case log, with second-surveyor sign-off for junior staff.
That record supports supervision, but it is not the reliability decision itself. The standard requires the surveyor to set that decision out in writing: the assumptions relied on, any concerns about the output and whether it can be used for its intended purpose.
Ashley verifies everything against his photographs and what he saw on site. Damp is a typical example. The AI might identify and describe a defect but not convey its seriousness or give enough advice. That is where he steps in.
'It's a helpful tool, but not something to be relied on to do your job,' he says. 'If I can see it's wrong, it's my job to correct it.'
He has also seen AI-assisted content elsewhere stray beyond the agreed scope of an inspection. Only a surveyor who knows what was and wasn't inspected will catch that.
His concern is reports being sent straight to clients from site without proper review. 'As a profession, we should be reflecting on that.'
For Anthony, the boundary sits with the physical and interpretive work that defines surveying: intrusive investigation, understanding building pathology, and speaking with clients. As he puts it: 'technology can't replace a surveyor's nose for smelling gas'.
His caution about what goes into open AI models also reflects another strand of the RICS standard. Firms must have clear policies on data use and AI governance, and where AI has a material impact on the service, clients must be told in writing and in advance when and for what purpose it is used.
For practices using AI in this way that means revisiting their terms of engagement, which must set out when AI is used, where in the process and whether the client can opt out.
'Surveyors must assess the reliability of AI outputs and remain accountable for all their work'
What other practices can take from this
One practice featured in this article is a team, the other a single surveyor, yet their findings overlap closely.
- Start with the input: both Ashley and Anthony found that adopting AI changed how they inspect, not just how they write.
- Treat it as cultural change: 'start the conversation with your surveyors early, run pilots and understand how each person works,' says Anthony. 'Any new implementation takes time. I thought people would see the opportunity quicker.'
- Keep the surveyor in control: every output is checked against the evidence and the agreed scope before it reaches a client.
- Know your obligations: the RICS standard's requirements on governance, reliability and client communication are mandatory, so build adoption around them from the start.
Ashley recently spoke to a surveyor weighing up whether to use AI. His advice sums up both experiences.
'Try it and see. It's there to help you, not to do your job. As long as you don't rely on it, it's a very useful tool. The second you become reliant on it, you'll have problems.'
Ian Hughes is marketing manager at Surventrix
Contact Ian: Email
Ryan Oakes AssocRICS is head of sales at Surventrix
Contact Ryan: Email
Related competencies include: Construction technology, Data management, Ethics, Rules of Conduct and professionalism
Discover the new RICS Member App: CPD on the go
RICS has introduced a refreshed CPD approach that prioritises meaningful, high-quality learning that genuinely benefits your work and is tailored to your specialism, career stage, and the real-world challenges you face.
The new app makes logging CPD simpler and more intuitive, so you can focus on the development that matters to your practice.
Technology and AI Summit
30 November | 08:15–18:00 GMT | 6.5 hours CPD | London
How can technology drive smarter, faster and more responsible outcomes across the built environment?
Join the inaugural Technology and AI Summit, featuring industry leaders, practical case studies, interactive workshops and live technology demonstrations to help professionals move from experimentation to informed adoption.
Explore key topics, including:
- AI, digital twins and smart building technologies
- data-driven investment, valuation and asset performance
- responsible AI, governance, ethics and professional judgement
- geospatial intelligence, automation and digital surveying
- digital skills and innovation for the future of the profession
- sustainability, operational efficiency and scaling technology adoption.
The headline keynote speaker will be Professor Dame Wendy Hall DBE FRS FREng, global authority on AI, technology and digital policy and adviser to the UK government and the United Nations.