The professional ethics of AI in surveying

But the launch of generative AI software ChatGPT in 2022 threw the technology firmly into the global spotlight and it swiftly became part of everyday life for many individuals and organisations. 

In the workplace, the technology’s impact has been profound, helping to create business efficiency and better time management. An RICS report, Artificial intelligence in construction 2025, found that nearly half of RICS member companies use AI to some extent – although within that figure only 2.5% reported widespread adoption, compared with 34% who are in early pilot phases of AI implementation.

But there’s a storm brewing, one loaded with caution over the technology, how to use it, the job losses it could lead to and how it could make cyber-attacks easier. The growing power of AI, the way it is being implemented and its possible risks prompted RICS to launch its professional standard Responsible use of artificial intelligence in surveying practice, which came into effect in March. 

The standard represents a major shift in how professionals understand and apply AI in the surveying industry, while looking at common misconceptions about AI and its potential to transform the sector, along with the ethical issues surveyors should know about.

“The new standard normalises AI use, gives clarity in what are we using it for and reassures clients and the public we’re using it responsibly,” says Nella Pang MRICS, MD of Omega RE. Trust and integrity are what RICS stands for, says Pang, which means not letting AI tools “make their own assumptions. It is about how surveyors use the technology to our best ability to ensure we’re giving our clients the most credible information.”

Pang was speaking on the RICS Podcast with James Garner FRICS, head of AI and data at Gleeds, both of whom were part of the RICS expert working group that developed the AI standard. The episode was hosted by Chris Alder, senior executive officer for standards and regulations at RICS, who started with the simple question of defining AI as we know it. 

The Organisation for Economic Co-operation and Development’s formal explanation, says Alder, describes it as being a machine-based system that generates outputs, such as predictions, content, recommendations or decisions that could influence physical or virtual environments. 

But different AI systems vary in their levels of autonomy and adaptiveness after deployment. The new standard clarifies this by focusing on material use and its impact on surveyors’ results.

Generative and analytical are the two distinct AI branches, with the former popular for creating new material and the same prompt giving different answers. The latter is more focused on processing structured data to make predictions and identify patterns.

“The new standard [recognises] we’ve a baseline understanding of AI and the different flavours of AI and what they’re good for,” says Garner. He adds that if a surveyor is doing analytical tasks, they use an analytical algorithm, whereas for creative tasks or role playing with AI, generative is better. “They’ve both got their place. It is finding the correct tools for the job,” says Garner. 

When it comes to people, there are four different AI personas as defined by Reid Hoffman, LinkedIn’s co-founder, says Garner: doomers – the technology is an existential threat to humanity; gloomers – those worried about the impact of the tech on their lives; bloomers – who use it ‘responsibly’ and at a moderate pace; and zoomers – who have completely embraced AI. 

“I probably fall into that last camp,” admits Garner, who says there is no right or wrong and people can be those personas at various times, but it does matter when it comes to a business context. 

Organisations, whether big corporations or SMEs, need to define their stance on the technology first, he says, before addressing employees’ apprehensions. Worries about issues such as bias can stop AI being used. “The barriers are around mindset more than anything else,” says Garner. 

Culture is another factor, says Pang, as some grew up with this technology, whereas older generations can find it new and fast moving: “They’re nervous to even try it because they don’t know what the exposure is.”

Pang is keen to emphasise that AI in surveying is not about replacing surveyors with robots, but using tools to analyse data and detect patterns that help them in their work: “AI is not making the decisions. It needs a human being to make the professional judgement and that distinction matters.”

Having the human in the loop is crucial, says Garner, who acknowledges the technology is “a real opportunity” for surveyors, enhancing what can be done by enabling AI to take away “the grunt work” and let surveyors spend more time with clients. 

“Used properly, it’s incredibly powerful,” he says, “but humans must retain control.” Giving an analogy of bookends, Garner says the human should be on either side, with AI in the middle being constrained. “The human, the chartered surveyor, starts by taking responsibility for telling the AI what to do and what we expect of it, just as you might tell an apprentice or a graduate what to do.”

Ultimately, grasping what the AI produces is essential, so results can be clearly explained when a client requests information about them.

For protection, Pang recommends two-factor authentication, avoiding similar passwords, having a separate phone for banking, updating privacy policies and GDPR, revising employee handbooks and providing anti-money laundering training for staff.

If a data breach happens, it should be reported to the Information Commissioner’s Office within 48 hours, while looking at the consequences for clients. “It’s about managing risk,” says Pang.

An additional concern comes from the possibility that organisations and individuals may not realise one AI tool is feeding data to another AI tool. “Understand whether they are linked,” says Pang. “You’re part of this one company, but your parent company is also over here. Are there connecting companies? One may sit in the likes of China – do they have access to that data?”

AI can also surface underlying data issues – not only with passwords, but with content management and file-sharing systems such as SharePoint. “My advice is not to go too fast if you’re implementing more complex uses of AI, where you are linking it up to your company’s SharePoint,” says Garner. 

And a wary eye should be kept on free models such as ChatGPT or Gemini, with Garner warning: “Nothing is free in this world. There is a payoff.”

The reason they’re free is because you’re giving them permission to basically utilise that data… which is a huge concern when you’re dealing with confidential data.”

Despite the downsides, Garner says AI needs to be seen as an opportunity where people have been given a powerful tool and they should not be scared about it: “Be cautious but have no fear.”

Pang again reiterates AI will not replace surveyors. However, those who choose to ignore how to implement AI and govern it responsibly, “are the ones that will be replaced. There is no hiding from this. It’s coming, and it’s a question of how we adapt and continue to improve the profession.”

This article first appeared in the June 2026 print issue of Modus and was based on a podcast recorded that February. AI has moved quickly since then. Recent weeks alone have brought the Hugging Face hack and warnings from a researcher at the AI company Anthropic about the dangers of how fast the technology is advancing.