The real AI challenge is not what you think
Surveyors UK
- Technology & AI
If you scroll through LinkedIn at the moment, you would think many have cracked AI implementation.
There are screenshots of clever prompts. Shiny demos. New tools launched.
In reality, most people I speak to are still juggling spreadsheets, email chains, PDFs and project files that live in six different places. They are being asked about an AI strategy while still fighting to find the latest version of a report. The real challenge is that most of us do not yet understand it well enough to use it safely.
Recent research found that the majority of professionals only have a basic to moderate understanding of AI, and almost half say lack of training is the main blocker to adoption. At the same time, firms are already buying tools for project management, cost estimation, risk assessment and more.
That is the wrong way round.
We skipped the boring bit
There is a pattern I see again and again.
- A firm buys an “AI powered” platform
- A handful of people are shown a demo
- Everyone is told to “have a play” and “see what it can do”
- Responsibility gets passed to IT
There is no shared understanding of what the model was trained on, how it fails, or where it must never be used without human checks.
In that world, it is very easy for things to look productive:
The danger is not the tool itself. The danger is a team that has never been given the time or training to understand what sits underneath. That is a knowledge problem, not a technology problem. And you cannot fix a knowledge problem just by buying better software.
What this looks like in surveying
If you work in surveying, you may already recognise some of this.
- Residential surveyors trying AI tools to tidy up report wording, without being clear what information they are allowed to paste in
- Quantity surveyors leaning on AI to check tender documents, with no process to validate what comes back
- Building surveyors asking AI to “list likely defects” from a vague brief, then feeling uneasy because the answer sounds right but does not quite line up with their instinct
None of this means surveyors should stop using AI, but it must be used responsibly
Shadow AI is already inside your business
Even if your firm has not rolled out any official tools, AI is already in the building.
People are using ChatGPT, Claude and Gemini on their own devices to help them in their work. This “shadow AI” is understandable. People are under pressure and the tools are easy to reach for and enticing.
The problem is what happens when this is done with no policy, no training and no governance. In many firms, people are already using generative AI on live projects with no oversight, producing outputs that look solid but have fundamental gaps or errors.
If nobody knows this is happening, nobody is checking the risks until something goes wrong.
Start with knowledge, not tools
So what should surveying and construction firms be doing instead
I think the starting point is simple. You build understanding first, then you choose tools.
In practice, that means four things.
AI literacy for everyone, not just “the AI person”
Treat AI like health and safety or professional ethics. It is not something one “AI champion” can deal with in the background while everyone else carries on as normal.
Everyone in the business needs a basic, shared understanding of:
- How these models work in simple terms
- Where they are strong, and where they fail
- When they are a useful assistant, and when they are a risk
That means:
- Giving people a plain English explanation of how AI tools actually work
- Showing real examples of “plausible but wrong” answers in a surveying context, so they see the limits as well as the benefits
- Making it normal to question AI outputs, instead of feeling embarrassed to say “I am not sure I trust this”
This is exactly why I ran a live two hour AI in surveying session on 28th January with AI trainer, Nick Crawford.
Nick brings the technical depth. I bring the surveying and business context. The aim is not to turn anyone into a data scientist. It is to get you and your team to a point where AI is something you understand and can have an informed conversation about, not something mysterious that is happening to you.
Map where AI actually helps
Not everything should be automated. Start by asking:
- Where are we wasting time on repetitive, text heavy tasks
- Where do we need faster drafts, but still want human judgment on the final decision
- Where would an AI co pilot help a junior surveyor, without replacing the need for senior review
Too many firms buy tools first, then try to work out what to use them for later. A basic process audit is far more valuable than yet another software demo.
Put basic guardrails in place
This does not need to be a fifty page policy.
- Be clear what data can and cannot be put into public AI tools
- Decide which parts of the workflow require mandatory human sign off
- Write down what good practice looks like for prompts, checking answers and documenting where AI has been used
The point is not to shut everything down. It is to make sure people know how to use AI in a way that protects clients, the business and themselves.
Ask the right questions with vendors
If you are buying “AI inside” software from others, ask direct questions:
- What was it trained on
- Where does it work well, and where does it fail
- How often is it tested or audited for bias and accuracy
If a vendor cannot answer those questions, your problem is not AI. It is transparency.
Why this matters for project delivery
Surveyors and construction professionals are trusted because they understand risk and take care with their judgments.
If AI is dropped into that environment without knowledge, it flips the model. Decisions start to lean on systems nobody really understands, and by the time problems show up in a project, it is expensive to untangle what went wrong.
A simple choice
AI is not going away. The tools will keep getting cheaper and more powerful. Your clients will expect you to understand them.
The real differentiator will not be who has access to a chatbot. It will be who has taken the time to build genuine AI competence across their team.
You can either:
- Invest in understanding now, then deploy tools in a controlled, confident way
- Or roll tools out first, and pay for the clean up later
If your name is going on the report, the letter, the valuation or the advice, that choice matters.
Continue the conversation inside The Surveying Room
If you prefer to learn in conversation, you are very welcome to join The Surveying Room, the free community space I run under Surveyors UK.
Inside we have:
- An AI and Tech discussion area where surveyors share what they are trying, what is working and what is not
- Space to ask questions you might not want to put on public social media
- Conversations across residential, commercial, QS, land and more, so you can see how others are approaching the same challenges
It is free to join, independent of any professional body, and open to anyone in or around the surveying profession in the UK.
I invite you to join the hundreds of surveyors already in The Surveying Room – register on Surveyors UK today and join the first independent community for the UK surveying profession.
Nina Young
Surveyors UK