From pilot to practice: AI in the built environment

Among respondents in the construction industry, that figure is 37%, pointing to widespread AI adoption across the built environment. In commercial property, just 20% of companies reported no AI usage at all and 29% said the same in construction.

These numbers come from RICS’ AI in commercial property and construction report 2026, which combines the results of the Global Commercial Property Monitor and Global Construction Monitor.

Tom Storey MRICS, one of the authors of the report says it shows that AI use is now mainstream: “Around two-thirds of construction respondents and more than three-quarters in commercial property use it somewhere in their work.”

“Privacy and security has become the fastest-rising concern in construction,” he adds. “Starting is no longer the hard part – turning a pilot into governed, routine work is.” 

The data below shows that commercial property is further ahead in making use of AI than construction, although in both sectors 39% of respondents reported the implementation of early-stage pilot projects.

Storey says most pilots begin with AI that arrives inside software the firm already pays for: drafting, summarising a document set, creating meeting notes. 

“In construction, that means a cost manager testing AI on tender document checks, or the estimating team using it to collate tender submissions, or a building surveyor trialling photo-based defect tagging,” he says. 

“In commercial property, the tools are built into portfolio, valuation and asset analysis software, so pilots are lease abstraction, comparable evidence searches and first drafts of valuation reports. The common pattern is one task, one team, and a person checking every output.”

In construction, the top reason given as a barrier to AI construction is a 'lack of skilled personnel’. In commercial property that was also a significant barrier but the most cited barrier to adoption is 'data quality and availability’.

Both sectors also named ‘integration with existing systems’ as one of the main barriers. 

In the US, a survey on AI adoption by Commercial Observer found that 44% of firms said their investment committees distrust AI-generated analysis, while only 27% trust the technology for financial underwriting. So, although the majority of commercial property companies are involving AI in their work to some extent, concerns over its reliability remain prevalent.

As well as being further along in terms of AI adoption, commercial property companies are also more likely to be increasing their investment in it over the coming year. Only around 20% in each sector said they had no plans at all for AI investment.

Storey thinks there are a few reasons for commercial property companies being more inclined to invest in AI. “First, more in commercial property report regular use of AI – 29% against 19% in construction – and investment follows proof. Second, the data is different. A [property] portfolio's data carries over from one asset to the next, so a tool that works on one building works on the next. 

“Construction data resets at every project and is produced by many firms to different standards. My own view is that construction buys for the project in front of it, so a capability that pays back across [multiple] projects is a harder internal case to make.”

One of the main aims of the AI standard is to help RICS members and regulated firms adapt to the use of this new technology by establishing guardrails to maintain professional judgement. 

Storey says RICS is working to increase the reach of the standard, for example through CPD, including a webinar on the report. “There are also a series of articles in Property Journal, beginning with AI procurement, the Technology and AI Summit on 30th November, and the launch of the Tech Network, which connects members with the solution providers whose products they will need to assess against the standard,” he adds.