Executive Summary: Key Takeaways
Commercial real estate AI pilots have jumped from 5% of firms in 2023 to 88% of firms in 2026. Despite this, more than 60% lack the strategic, organizational, and technical readiness to move those pilots into lasting business impact.
The question that surfaced most clearly on June 11th was what kind of organization you need to become for AI to actually matter. Strategy, not tools, is where that gets answered.
Unprecedented cybersecurity risks abound. Intended and unintended outcomes are rising at an equal pace. Without a detailed plan, organizational readiness, and proper governance, the risks of getting it wrong are profound.
In the days ahead, we will share clips from the event, and we will be back in September with our next speaker event. If you were not in the room on June 11th, we hope to see you then.
Evening
On June 11th at Mabel Chelsea, MaryAnne Gilmartin hosted myself and Jack Hanlon for an intimate conversation with a curated group of real estate executives centred around a timely, forward-leaning conversation about AI. The dizzying pace of change in AI technologies leaves everyone in the industry grappling with a very challenging organizational question: what will it take to make AI work at scale in my company, and commercial real estate more broadly?
MaryAnne’s framing of why the event existed said more than any agenda could.
“This conversation was born of many discussions inside MAG Partners about how confused, frustrated, and aghast we are about the state of AI in our industry.”
MaryAnne Gilmartin, CEO, MAG Partners
The audience included leading developers, investors, lawyers, architects, engineers, and advisors who have been taking the AI question seriously and who came with deep questions, a quest for clarity, and a desire for direction. It was an engaging, fun, and enlightening conversation that could easily have extended for another hour. It was great to be a part of it. The event was heavily oversubscribed; as Roy Scheider famously said in Jaws, "You're gonna need a bigger boat". So, for those of you who could not be in the room, in this piece, we have captured the key insights and takeaways to share with you. Hope to see you when we get together again in September.
The Data: Where AI Stands in Commercial Real Estate Right Now
The JLL Global Real Estate Technology Survey, which came up in our discussion, drew responses from more than 500 senior decision-makers across 15 markets. The data shows AI piloting in commercial real estate has jumped from 5% of organizations in 2023 to 88% today, with the average organization now running up to five AI pilots simultaneously. 87% of firms are increasing their technology budgets because of AI, with the top five spending priorities all AI-related.
The same research, however, identifies the problem running alongside all of that activity. More than 60% of organizations lack the strategic, organizational, and technical readiness to take their pilots to scale. Most companies spending money and energy on AI right now are doing so without the foundations to make it durable, sustainable and scalable.
JLL’s findings create a clear image of the commercial real estate industry we are seeing in the field when I talk with clients in the industry daily. One where people are eager to experiment but unsure how to create a sustainable business impact.
Jack opened the evening by reaching back to a pattern economists identified in the 1980s. Robert Solow famously noted that computers were everywhere except in the productivity statistics. What his research actually showed was dramatic variation at the firm level. Some organizations saw exceptional gains. Others saw negative returns. Almost no one landed in the middle.
“People are either going to see significant productivity gains or they will see potentially negative productivity. Very few people are going to see the middle.”
Jack Hanlon, Leading AI Executive and NYU Stern Professor
That pattern is playing out again with AI. I have led successful transformations at some of the world’s largest enterprises and built teams and products from scratch throughout, but two things are new this time; the pace of advancing technology we are seeing today is unprecedented, and the gap between the winners and the rest of the pack is widening quickly.
What AI Actually Looks Like in Practice
Not surprisingly, the term "Using AI" covers a wide range in practice. At one end of the spectrum, individuals are moving faster on personal tasks using LLMs like ChatGPT, Claude, and Gemini. On the other end, Jack described how in the minutes leading up to this event kick off, he unleashed multiple research and writing agents in parallel, synthesizing their outputs into a finished document in his own voice, and sending it to his team while guests were still arriving. Between those points sits everything from AI-assisted business development, to automated competitive intelligence, to compliance documentation. Work that used to take multiple people days to complete, is now done in minutes or hours.
There is a real difference between individuals using AI tools for personal productivity and organizations that are AI-enabled, where AI runs through how every function is approached. While personal productivity gains are great, transforming an entire team or company has dramatically larger impact: every team operating with more speed, more precision, and better information:
“As a fully AI-enabled company, maybe you could double in size without hiring any more people.”
David Dunne, CEO, Elemental AI
MaryAnne’s account of MAG Partners' own path with Elemental AI as their trusted advisor is a useful illustration of what the early, deliberate phase looks like in practice. The firm is using AI to compete for more business with less time and resources, producing stronger proposals, and removing redundant tasks that drain organizational bandwidth. A deliberate start, with clear priorities and guardrails that protect what matters.
The Pointy Tip Of The AI Spear
The conversation went to some of the big challenges facing firms in their pursuit of AI. Three of them stood out:
The right ROI analysis. Most AI investments are being evaluated against the wrong benchmarks. Companies focused on efficiency and cost reduction are measuring a secondary outcome while the real competitive race runs on something else: market intelligence, investment decision quality, customer retention, and the ability to identify opportunities before competitors do. The JLL data is explicit about this. The AI objectives most strongly correlated with high achievement among leading organizations are growth-oriented rather than efficiency-oriented. Getting the objectives right is a strategic decision, and most firms have not made it deliberately. You can run a busy AI program that performs well on efficiency metrics and still fall behind.
The cost of moving without a plan. When AI flows through a company without governance, unvetted output that looks credible at a glance floods decision-makers, creating an avalanche-esque burden rather than efficiency. On cybersecurity, the exposure is more immediate than most in the room anticipated. Walking through the mechanics of a recent Claude Code leak in which major threat actors had forked the exposed codebase within 24 hours, creating attack vectors against every user of Claude Code, the room was justifiably concerned about the consequences that could come from inadequate guardrails. I described a (former) CFO who found his team locked out of their digital bank. With no branch to visit, and a full takeover of their banking signature process, two million dollars was wired from the company's accounts to crypto accounts in a dozen countries by hackers. In minutes the funds were gone and unrecoverable. These are not hypothetical scenarios.
Shadow AI is a major concern. Across organizations of every size, staff members are already using AI tools the leadership team often has no knowledge of, and with company data that has not been protected. This is a current governance question, not a future one.
“If you create the wrong incentives, you’re going to create the wrong behavior. And then you get the wrong output.”
David Dunne, CEO, Elemental AI
The companies navigating this well treat AI transformation as a leadership problem. They come in with clarity about what they're trying to accomplish, build governance to match, and create conditions for results to build on each other rather than for activity to accumulate.
What a Real Strategy Requires
Among companies already achieving their AI objectives, the JLL research points to three consistent characteristics: unified strategies across business functions, dedicated roles driving AI beyond baseline IT, and a willingness to rethink the workflows that truly move the needle for the business. The organizations with these characteristics are achieving their AI objectives at significantly higher rates than the rest of the field, and that gap is growing.
The most useful analogy of the evening came from Jack. When factories first adopted electric motors to replace steam engines, most simply swapped the power source and changed nothing else. Productivity didn't improve. The gains came only when organizations redesigned the factory floor around what electric power actually made possible: different layouts, different workflows, a different operating logic entirely.
“The more we’re thinking about how to just do the same things we do today, but with AI, the more likely we’re not going to see any change.”
Jack Hanlon, Leading AI Executive and NYU Stern Professor
Most organizations are stuck on the same problem: they lack a proven roadmap, a trusted guide, and a team that has navigated a transformation like this before. The companies making real progress have people alongside them with the credibility to say no as readily as yes: operators who have run P&Ls and been accountable for results; leaders who understand the difference between a roadmap and a slide deck; transformation veterans who will tell you plainly what won't work before you've built it.
“If you’re just dabbling, you’ll end up with the wrong results. None of that is possible without a plan and a roadmap of where you’re going.”
David Dunne, CEO, Elemental AI
For commercial real estate leaders, this is fundamentally a navigation challenge. The technology is there. The question is whether the conditions are in place across the enterprise to use it well.
See You in September
The AI question in commercial real estate is largely past that "technology" stage. The question that matters now is how your organization is building the conditions to leverage AI to its fullest potential.
Many thanks to MaryAnne and the MAG Partners team for such a memorable evening and to Jack for shining a light on how we should all be thinking about the AI opportunity and risks ahead of us. And for everyone who attended, as well as those who could not, we hope you can join us again in September. Details will follow.
David Dunne
