AI in Real Estate – Nirvana or new bubble?

Everyone is talking about it – but is artificial intelligence really having an impact on how business is done in real estate?

To find out, real estate adviser Chris Dougray assembled a round table of key players in the sector including developers, operators, funders, and industry bodies at Shepherd and Wedderburn’s Edinburgh office to see if its effects were being felt in Scotland and the rest of the UK.

US real estate expert Brad Hargreaves joined as keynote speaker via Teams to share his experience of how AI is being deployed across the Atlantic.

The discussion was held under the Chatham House rule, but here are the key themes that emerged from a very lively debate.

US operators are leading the way

Use of AI within real estate has grown rapidly in the past two to three years – particularly within Built to Rent and student living projects.

Yes, there is a lot of hype, but there are also now tangible efficiencies, savings, and NOI (Net Operating Income) improvements being generated on the operating side.

A recent study in the US showed that Build to Rent operations (known as Multifamily in the US) are near the top of the AI adoption curve in terms of how aggressively real estate operators have embraced its use.

It was clear from the feedback in the room that the UK real estate sector was well behind that curve compared to the US. A worry was expressed that politicians and regulators in Scotland and the rest of the UK are not focusing enough attention on how the economy will be impacted, or how it could benefit if the right conditions can be created for new technologies and services to be built on this side of the Atlantic in order to grab a slice of the potential economic benefits.

What is AI being used for in Real Estate?

Leasing is a specific area where unstructured conversations have historically taken place with leasing agents at front desks – someone reaches out with an inquiry about a unit, they have some back and forth, and they schedule a time to take a tour. Now, all of those processes are being done by AI.

And, according to feedback and KPIs, it does a better job than humans in terms of lead to lease conversion.

Another perhaps more surprising example of successful AI use on the operations side is managing payment delinquencies and collections.

Recent case studies have shown that when those processes are run by AI there are not only efficiencies of scale, but tenants respond well to it – meaning cash collection improves significantly.

Managed office operators have also started watching what Build to Rent is doing very closely and using the same automated tools to run Grade A small suites much more easily.

In development and construction there is a lot of interest in AI and there are a few areas where it’s being tested in practice, but it’s at a much earlier stage than in operations.

However, it is really being deployed on the front-end site assessment, including underwriting, and sourcing planning documents – and it’s changing the pace these tasks can be done at.

Instead of being able to assess two or three sites a week, developers are using AI to search for and grade tens of thousands of sites very quickly to create shortlists in line with their own criteria. These are then handed over to experienced humans for final assessment to see if they stack up.

AI is being used to make everything from site sourcing all the way to making an offer on a piece of land much more efficient – and they can go through many, many more sites in a given period of time with the same or less bandwidth.

The data centre conundrum

Another key issue discussed at the roundtable was the expansion of a key resource needed to power AI in real estate – data centres.

The demand for power to run them is already creating a strain on interconnection capacity and the electric grid system in the US. That in turn is driving interest and investment in technology such as micro grids, small modular reactors, and reform on permitting transmission lines.

However, there has also been community pushback on data centres due to concerns about energy costs, environmental impacts, and the fact they don’t bring with them significant numbers of local jobs compared to a factory or an office occupying a similar site.

There is a question mark in the US over planning permissions for data centres and a feeling that unless an answer is found to the community benefit question the window for building major new ones may close rapidly.

Similar issues are already arising in the UK and Europe, but there were concerns expressed around the table that at a global level, data centres could not be entirely outsourced due to infrastructure security concerns – so a homegrown solution for each locality is needed.

Who benefits from AI in real estate?

Some present suggested that AI increasing productivity is a net positive for society at large.

However, for real estate specifically there was no consensus from the businesses around the table that they would feel a direct benefit in the immediate future.

In fact, there were concerns raised that it could have a negative impact on jobs, pricing, and legal liabilities.

It was highlighted that in US venture finance circles there is a debate on where AI gains are going to land. Will it be with end users, or just big tech companies?

And are real estate business owners really going to be willing to share their confidential commercial data with outside AI software from a risk perspective – or if it could be used to train an AI model that their competitors will also use to benefit their companies?

These are questions for which there may be no clear answer just yet, or until AI use matures within the real estate sector.

Next steps for AI in real estate?

Thank you to Chris Dougray and Brad Hargreaves for facilitating a very open and thought-provoking debate, and to the valued clients and guests of the firm who joined the roundtable.

If you’d like to understand more on the commercial and legal implications of deploying AI in real estate within the UK, get in touch with Shepherd and Wedderburn’s market-leading Real Estate and Infrastructure team to arrange a briefing.

Contributors: