Artificial intelligence is becoming part of everyday building operations, from predictive maintenance and energy optimisation to service desks and occupancy analytics. As adoption grows, facility managers face a new responsibility: governing how AI is used, not simply what it can do.
AI is increasingly embedded in CAFM (computer-aided facility management) platforms, building automation, energy management and service desk software. It can identify unusual consumption, predict maintenance needs, analyse occupancy and support operational decisions.
As highlighted by Facility Management.de, the first governance challenge is therefore visibility. FM teams need to know where AI is being used, what data it processes, which decisions it influences and who remains accountable.
This is becoming more important as the EU AI Act takes effect. Certain AI applications are subject to transparency requirements, while some uses face tighter restrictions.
The algorithm is not responsible
AI can recommend an intervention, but operational responsibility remains with people.
A predictive maintenance system, for example, may detect an anomaly, but statutory inspections, manufacturer requirements and contractual obligations still apply. Similarly, AI can optimise heating, cooling and ventilation, but energy savings cannot automatically take priority over safety, hygiene, indoor environmental quality or occupant comfort.
For facility management teams, AI governance is therefore becoming an operational issue rather than simply an IT concern.
Put AI into contracts
Governance also extends to suppliers. Operators should understand which AI components their suppliers use, what building and user data they process and whether external AI models are involved.
Tender documents and FM contracts may consequently need to address AI explicitly, covering data use, documentation, model updates, interfaces and human oversight.
Start with an AI inventory
Kay Meyer, Managing Director of FM-Connect.com Network GmbH recommends starting with a company-wide AI inventory. A sensible first step is straightforward: identify what is already in use.
An AI inventory should include not only standalone tools, but AI functionality embedded in CAFM, automation, energy, security and workplace platforms. For each system, FM teams should understand its purpose, data sources, supplier, users and internal owner.
AI offers substantial opportunities for facility and property management, but it also changes the operator’s role. The key questions are increasingly simple: Where are we using AI? What decisions does it influence? And who is responsible?
The technology may be intelligent. Accountability still belongs to people.