Why Facilities Teams Risk Operational Disruption If AI Arrives Before Readiness

The appetite for AI in facilities management is real. The readiness is not. While organizations across the sector talk openly about integrating AI into operations, the foundational work required to make those tools useful is largely unfinished. Asset histories are incomplete, process data is siloed, and the systems that track building performance are rarely linked to one another. The result is an industry where AI remains mostly on paper, and the teams rushing to implement it without addressing those gaps risk making operations worse, not better.

Dory Antoun is a Certified Facility Manager and PMP-certified project management professional with over 12 years of experience managing complex FM operations in Saudi Arabia. He is a certified member of both IFMA and IWFM, and writes regularly on FM operations and data maturity. From his perspective, the conversation about AI in facilities management is getting ahead of itself.

"Facilities management and AI are still mostly on paper. The appetite is there, but if you want to bring AI into the business, you have to be well prepared with proper structure and proper data in place. Otherwise, it's going to jeopardize the operation," says Antoun. The gap Antoun identifies is not a technology gap. It is a data gap. Most FM organizations maintain some form of asset history, but the records are often incomplete, inconsistently formatted, and disconnected from one another. Without clean, linked data across systems, AI tools have nothing reliable to work with.

• Clean house first: Antoun frames the prerequisite in direct terms. "You have to clean your house first. Put your house in order, then you can bring AI, and you will benefit. Whether it's operations, hard services, soft services, commercial budgeting, or resource planning, you can have a very high return. But before doing that, you have to know how to feed it."

• The shortcut trap: Without that foundation, most organizations default to the path of least resistance. "Most organizations just bring AI in and end up automating minor tasks instead of benefiting from its real potential," Antoun says. Instead of targeting high-value operational use cases like predictive maintenance or resource optimization, teams settle for digitizing low-impact admin work and call it transformation.

The higher-value target, in Antoun's view, is predictive maintenance. In FM, condition-based and predictive approaches have historically relied on manual measurement, where a technician physically inspects equipment and estimates remaining useful life. AI can automate that process, but only when years of properly structured asset performance data are available to train on.

• What good looks like: For Antoun, the benchmark is measurable. "What I see as good is minimally 50 to 60% less downtime on critical equipment. That's where AI can create real value in the FM business, if properly implemented."

• The real cost: The financial argument for doing the preparatory work is equally clear. "It takes a lot of money if you don't have a proper system in place. And it takes zero money if you build it up with time," Antoun explains. Teams that follow a structured process and invest in data maturity over time avoid paying twice: once for a failed implementation and again for the remediation. Those that skip ahead end up spending more to fix the disruption than the AI would have saved.

But the most significant barrier Antoun raises is not organizational. It is structural. Industry governing bodies, the very institutions that facilities professionals rely on for standardization and professional development, have yet to produce meaningful AI guidance.

• The missing framework: "IFMA and IWFM don't have proper AI material in place yet to educate the subject matter experts toward implementing AI," Antoun says. "This is the missing piece. Without standardization from governing bodies, companies are left guessing on how to plan and implement AI properly."

Without that shared framework, even well-intentioned organizations are left improvising their own approaches, producing inconsistent results that cannot scale across the industry. In Antoun's view, standardized guidance would give FM teams a common playbook for assessing readiness, structuring data, and sequencing implementation, the same kind of rigor that already exists for areas like fire protection and building compliance.

Until that guidance arrives, Antoun's advice is to resist the pressure to implement AI for the sake of having it. The preparation phase is not a delay. It is the work. "This is the missing link," Antoun concludes. "You have to do a pre-implementation exercise. Check the existing systems, decide whether it's beneficial to bring AI now, or whether you need a few months of restructuring first. That step is what separates a real return from jeopardizing your operation."