Are AI tools good investment drivers?
Almost 70% of JLL’s sales efforts now see potential buyers identified through artificial intelligence (AI).
From a standing start, the real estate services company has seen its AI-first approach to identifying buyers explode in the last year, according to Jorge Martín, its global head of innovation models, capital markets. A survey published by the company last year found that 88% of all real estate investors were piloting the technology and, of those, 29% were trying to use it for acquisitions and dispositions.
In December, a real estate financial modelling platform called Built AI announced it had secured $6m (£4.4m) in seed funding to support its growth in the US.
The company says its platform can instantly extract and analyse building data to help clients underwrite new investment opportunities. It adds that it can create full business plans and scenarios for any asset, in a matter of minutes.
So, can AI really help spot investment opportunities better than human beings?
Nimbus is a widely used platform that helps property professionals research and spot opportunities. The company’s chief technology officer Nic Neate says one of AI’s “most immediate strengths” is its ability to process and interrogate large volumes of data at speed.
Lisa Talia Moretti, a digital sociologist and co-chair of the British Interactive Media Association’s AI Council, agrees that data analysis can be very swift but says investors should be wary of relying on it too much.
"The future is unknown to all of us, no matter what the tech bros would like us to believe"
Lisa Talia Moretti, AI Council
“AI is built with real-time or past data, and while that can be helpful to model and imagine a possible future, it shouldn’t be taken as fact,” she says. “The future is unknown to all of us, no matter what the tech bros would like us to believe.”
Neate stresses that identifying investment opportunities is a “nuanced process” whereby AI can aid human judgement and experience.
“When used effectively, AI can significantly enhance a human-led search process, enabling investors to identify patterns, anomalies and potential opportunities far more quickly than through traditional manual methods,” he argues.
Over at JLL, Martín estimates its own system helps sales teams bring in bidders that would otherwise not have been approached in at least one in 10 cases.
Research on who would be the likeliest buyer is now carried out by its AI system in 68% of attempted global transactions, he says.
“It’s expanding the capabilities of our people, because before it used to take hours or even a couple of days to gather the list of people who might want to take up a buying opportunity. Now, AI is doing that in seconds.
“We have various smart people and they know a lot about the market, but […] 10% to 15% of the time, AI is bringing newcomers into markets or sectors – [people] the broker wouldn’t have called because it was not obvious.”
The capacity to connect different pieces of information in real time was impossible before the AI system was introduced, he says.
Humans still centre stage
JLL has brought in a combination of generative and traditional AI across both its customer management and opportunity-spotting platforms and has spent a lot of time getting its data in order, says Martín, adding that the AI also reads company communications to aid its efforts.
But humans need to remain at the centre of a process that is designed to prompt them.
Martín adds: “It’s a facilitator. It could speed things up; it could bring you to a market that you didn’t expect. But there is still a lot to be done. And the most important thing to do is close a deal – reach the owner and connect with the individual. There is a lot of relationship between humans, especially in the last mile. Deal-making is still very human.”
Neate agrees that people need to remain at the heart of processes. “While advances in agent-based systems and predictive modelling are accelerating rapidly, investment decision-making still benefits greatly from contextual understanding, market intuition
and strategic insight – areas where human expertise remains critical,” he says.
"AI is a powerful tool, but it must be applied with discipline and oversight"
Nic Neate, Nimbus
MartÍn is extremely enthusiastic about the use of AI and spends much of his time training teams in how to get the most out of it and then updating the process when the technology improves.
But as part of his work, he says, he needs to stress the limitations of the tools. “People sometimes think it’s magic and perfect, so they get frustrated if they get a result that is wrong,” he explains. “AI is complex; it has ‘hallucinations’ and sometimes it is not right.
“I have to explain that it is probabilistic; it is giving you ideas. It could be wrong and sometimes you don’t know why because of how AI works. It’s not going to be perfect.”
A lesson learned at JLL so far during its experience with AI is not to blindly accept results – and not expecting external clients to either.
Martín and his colleagues also work to demonstrate how the results translate into understandable labels for the business and try to demonstrate the underlying reasons for a suggestion where they can.
“If people see that underlying reason and understand it, it can resonate with them,” he says. “Sometimes, what happened with the pilots at the beginning of the journey [is] someone opened Horizon [JLL’s opportunity scanning platform] and said, ‘oh look, AI is telling me that you are going to sell this building’ – without any context.”
Responsible use
Not questioning or understanding the outputs of the system, even if they are correct, makes it difficult to communicate with potential clients, Martín believes. And this can lead to a lack of trust among users of the tool and in the perception of the company using it.
The sentiment is echoed by Nimbus’s Neate. “Responsible adoption is essential,” he says. “The focus should be on building and deploying robust, well-tested systems that deliver reliable, transparent outputs.”
He adds: “AI is a powerful tool, but it must be applied with discipline and oversight, rather than relying uncritically on automated outputs.”
Before deploying the systems, however, one of the most important things is for a human being to decide exactly what they should be used for, according to Martín.
“Some of the AI tools are expensive,” he says. “Pick the right problem and make a cost analysis if that’s going to be really impactful, because sometimes even if you solve a problem […] a cost analysis [shows] it would have been cheaper with people than AI.
“Many times, people don’t think about it. It could become exponentially expensive if you pick the wrong problem and don’t have a good outcome, or [you have a] good outcome but it is not so relevant to your business.”
Humans, then, still appear to be key to successful financial decision-making in property. For the moment, at least.
Author: Ian Weinfass
Commentary: Nic Neate, CTO, Nimbus
This article originally appeared on Property Week - 25th February 2026
