Facilities management sits underneath almost every AI use case in NHS estates, government buildings, defence establishments, schools, real estate portfolios and industrial sites. The IWMS and CAFM platforms supporting FM are extraordinarily capable. They are also not going to govern an operator's AI for it. The FM sector's next five years will be decided by asset data discipline, not platform selection.

The sector below every other sector

Facilities management is the sector underneath every other sector in this series. Trusts, councils, departments, defence establishments, manufacturing plants and real estate portfolios all rely on FM to keep the estate operational. PPM schedules, reactive repairs, helpdesk, space management, cleaning, catering, energy, security, grounds, lifts. Hundreds of service lines, often layered across multiple suppliers on different contract cycles.

The AI ambition aimed at this estate is wide. Predictive maintenance on building plant. Space and occupancy optimisation. Energy and net-zero reporting. Repairs prioritisation. Automated helpdesk triage. All of it sits on the FM data record: the IWMS, the CAFM, the asset register, the supplier platforms and the occupancy systems.

That record, in most portfolios, is the least integrated part of the operation.

The IWMS and CAFM estate is capable but not governed

Modern IWMS and CAFM platforms (Maximo, Planon, Concerto, FSI Concept Evolution, MRI Concept, Archibus, Nuvolo, and others) are mature and capable. Most operators have invested meaningfully in them. The limits on AI value, however, sit above the platform layer. They sit in governance.

A large FM operator typically manages data across several tenants, often one per client account. Each tenant has its own classification, its own asset hierarchy, its own criticality logic, and its own history of configuration decisions made by contract managers long since moved on. Consolidation of insight across that estate is where AI is expected to add value, and where data fragmentation most frequently blocks it.

This is not a platform problem. It is a governance problem. ISO 55000 provides the discipline.

The Account Director reads this as a bid-and-renewal defensibility problem. The Head of Estates reads it as an asset-level energy performance problem. The Head of Technology reads it as an IWMS-and-CAFM reconciliation problem. The next Scope 3 or net-zero reporting cycle will ask all three the same question.

Three areas where the data record decides the outcome

Predictive maintenance on building plant. HVAC, chillers, lifts, BMS, fire systems. All have mature sensor data. Many operators run predictive models on rotating plant already. Scaling those models across a multi-site portfolio still depends on the FM asset record recognising the same asset, in the same way, at every site. Most do not, yet.

Energy, net zero and ESG reporting. Scope 1, 2 and 3 emissions reporting is a board-level requirement for most corporate occupiers and most public landlords. Energy and carbon data flows through the FM record. AI can help identify optimisation opportunities, but only when asset-level energy intensity is linked, consistently, to the asset record. Where it is not, AI amplifies the noise instead of reducing it.

Helpdesk, triage and workflow. Generative AI is already being deployed on helpdesk triage, automating classification, routing and response drafting. The value is real. The risk is that an AI triage engine trained on inconsistent historical ticket data will confidently perpetuate inconsistent classification forever. The asset data governance question comes first.

The PFI legacy problem

Large parts of the UK public estate are still encumbered by PFI and PF2 contracts. The asset data held by the FM provider, the SPV, the public sector client and the sub-contractor often diverges. Contract expiries are approaching for many of these deals. At each expiry, the question of whose data is whose becomes live.

AI is arriving just as those handovers begin. The operators that have a governed, platform-agnostic asset record are well placed to absorb incoming estates cleanly. The operators that do not will be paying consultants to reconcile data at the worst possible time, when a contract hand-back is already in progress.

Supplier and cyber risk sits on the same record

The IBM Cost of a Data Breach Report 2025 found that sixty-three per cent of breached organisations had no AI governance policy in place. The UK Cyber Security Breaches Survey 2025 found that only fourteen per cent of UK businesses reviewed the cyber risk of their immediate suppliers in the last year. FM, by nature a layered supplier ecosystem, is structurally exposed to both.

The Cyber Security and Resilience Bill, progressing through Parliament in the 2025-26 session, will put supply chain cyber obligations on board agendas in 2026. FM directors who can evidence that their AI-linked asset data is governed, classified and access-controlled will pass board scrutiny. Those who cannot will be flagged.

The continuous layer: Data Governance as a Service

FM data does not become defensible once and stay defensible. Client tenants change. Contracts hand back. Suppliers rotate. Platforms upgrade. Classification logic erodes. Scope 3 and net-zero reporting cycles are annual; AI programmes built on the FM record are continuous.

Data Governance as a Service (DGaaS) is Brainwave Asset Intelligence's cross-sector model for that continuous layer: practitioner-led governance that detects duplicates, degradation and supply-chain integration gaps in the FM record across tenants, hand-backs and supplier changes rather than at them. The full treatment, applied consistently across asset-intensive sectors, sits in the DGaaS anchor on this page.

Foundations before automation

The AI use cases aimed at the FM record are real and the value on offer is substantial. The constraint is not platform choice or vendor availability. It is whether the asset record underneath, across tenants and across contracts, is complete, governed and defensible.

Data readiness is the foundation the next five years of FM AI work will be built on. Foundations before automation.

Key takeaways

  1. FM sits underneath the AI use cases in every other sector, and the FM record is the least-integrated layer in most portfolios.
  2. Modern IWMS and CAFM platforms are capable. They are not a substitute for governance.
  3. Predictive maintenance, energy reporting and helpdesk AI all depend on a consistent, governed asset record.
  4. PFI contract hand-backs will expose asset data problems. Operators who prepare now will absorb hand-backs cleanly.
  5. Supplier and cyber obligations under the Cyber Security and Resilience Bill will reach FM board agendas in 2026. The governance evidence a director can point to will come from the asset record.

Sources: IBM Security, Cost of a Data Breach Report 2025. UK Government, Cyber Security Breaches Survey 2025. UK Parliament, Cyber Security and Resilience Bill progress, 2025-26 session. ISO 55000 series. Verified 20 April 2026.