There is a version of AI in commercial real estate that changes how brokers work. And there is a version that puts the word “AI” on a press release and calls it done. Right now, both are being marketed with equal confidence.
Dan Mosher, CEO of DealGround, has a clear framework for telling them apart. His company built its platform from the ground up around AI, and a new joint study with First American Data & Analytics – the CRE Industry Pulse Check – gives him data to back up what he has been seeing in the field. Of the 255 CRE professionals surveyed, 34% say they cannot figure out which tools to actually use. That confusion is not accidental.
The Bolt-On Problem
When a legacy CRE platform announces an AI feature, the question worth asking is: did they rethink the product, or did they add something on top of it?
In most cases, the answer is the latter. An AI chatbot layered onto an existing database. A smarter search that mostly just autocompletes better. A filter that does what a regular filter used to do, but faster.
“A lot of the existing companies are bolting on AI solutions,” Mosher said. “Maybe they’re putting an AI chatbot on something, or they’re using AI to do a very specific function. But they’re not fundamentally changing their product.”
He describes this as an AI veneer: a thin layer of new technology applied to an old architecture, primarily because the market expects it. The underlying workflow remains unchanged. The broker is still jumping between systems. The data is still siloed. The output is, at best, marginally better.
What Native AI Actually Means
The distinction Mosher draws is between AI as an add-on and AI as a foundation. A product built AI-native means the original concept was designed around what AI makes possible, not adapted afterward to include it.
For DealGround, that means the entire data model, the search experience, the property research workflow, and the ownership intelligence layer were all built under the assumption that AI would end up executing the work end to end. A broker types a plain-language query, the system constructs it, runs it against a database of over 160 million properties, returns results that no public LLM could produce on its own and advances it to next steps not previously thought possible.
“It has to be materially different from everything that came before it,” Mosher said. “Using some of the latest innovations that AI has brought to bear to create step-function efficiency.”
That is a higher bar than what most current CRE AI announcements are actually clearing. The survey data supports this: even among the 66% of CRE professionals who use AI regularly, the top current use cases are market research, document abstraction, and marketing content. Nobody in the study named a platform that had fundamentally changed how they source deals or identify ownership opportunities.
Why the Incumbents Have an Incentive to Keep Brokers Confused
The confusion around AI in CRE is not entirely accidental. The established players, the ones with decades of market share and broker relationships, benefit when brokers stay on the fence.
If brokers wait for the dust to settle, they keep paying for the same subscriptions. If they feel like AI is already “covered” by their existing tools, they have no reason to look elsewhere. The bolt-on strategy is partly a product decision and partly a retention play.
“The incumbents that have been in business for decades are going to try to expand on the AI Doomer narrative,” Mosher said. “They want to create paralysis on the part of the brokers.”
The study’s own data reflects this. 31% of respondents report no measurable time savings from their generic AI tools. That is not a group that has switched to a better solution. It is a group stuck using something that is not working particularly well, unsure whether the problem is the tool or their use of it.
The Questions Worth Asking Before You Buy
If you are evaluating an AI-powered CRE platform right now, Mosher’s framework suggests a few things to probe.
Does the AI run against your private data, or just the public internet? A system that only queries public information is doing what a general LLM already does. The real value comes when AI works with the data that only you have access to: your own OMs, your rent rolls, your ownership research.
Does using it change how you work, or just how fast you do one step? A genuine AI-native product should collapse multiple systems into one process, not accelerate a single task while leaving everything else untouched.
Is the workflow end-to-end, or does it still hand off to something else? If you need four other tools to complete a deal process, the AI in question is handling one node, not the pipeline.
The wave of announcements makes it easy to assume the whole industry has moved. Most of it has not. Only 5% of CRE professionals in the survey say they are using an AI capability for sourcing and prospecting, one of the most valuable parts of the process. That gap between what is being marketed and what the market actually needs is the clearest sign of where real product work still needs to happen.
Download the full CRE Industry Pulse Check report here: http://report.dealground.com/cre-pulse-check-report-may-2026.
DealGround is the AI-native intelligence command center built for how brokers actually work – turning property, tenant, ownership, and market data into faster, more informed deal decisions. Learn more at dealground.com.
This article is based on information provided by the expert source cited above. It is intended for general informational purposes only and does not constitute legal, financial, or real estate advice. Readers should conduct their own research and consult qualified professionals before making any real estate or financial decisions.
Disclosure: Individuals or companies mentioned may have a commercial relationship with KeyCrew.
