EDITOR'S NOTE: We welcome new column contributor Michael Laurie to RENX. As the founder and president of Planit Engineering, he brings deep expertise in building measurement, BOMA standards, floor area certifications and digital building documentation.

Artificial intelligence is quickly becoming one of the biggest topics in commercial real estate.

Owners and managers are exploring AI for predictive maintenance, energy optimization, capital planning, due diligence, tenant experience and virtually every other aspect of building operations.

The opportunity is enormous.

But after spending decades measuring, documenting and creating digital records of existing buildings, I believe the industry may be overlooking a fundamental problem.

AI can only be as intelligent as the building information underneath it.

And for much of the existing commercial real estate market, that foundation simply isn't ready.

The AI ambition is ahead of the infrastructure

This isn't simply a technology problem.

BOMA Canada's 2026 Trends in Commercial Real Estate survey, conducted through its AI4CRE initiative, found only 10 per cent of respondents had fully implemented AI across their portfolios. Another 33 per cent were still in the early planning stages and 20 per cent reported no implementation.

Integration challenges with existing systems were identified as a barrier by 44 per cent of respondents. At the same time, more than 70 per cent expressed interest in smart-building automation, advanced analytics, predictive maintenance and sustainability monitoring and reporting.

That presents an interesting question.

Before we can apply sophisticated intelligence to a building, how well does the machine actually understand the building itself?

Buildings don't have a data shortage

Most commercial buildings already have an extraordinary amount of information associated with them.

There are drawings, leases, equipment schedules, BOMA calculations, condition assessments, photographs, maintenance records, energy bills, work orders, specifications, inspection reports, building automation systems and increasingly large quantities of sensor data.

The problem is that much of this information was created at different times, for different purposes, by different organizations.

Some of it is accurate. Some of it is outdated. Some exists only in PDFs or spreadsheets. Some sits inside specialized software. Some is stored on shared drives. And some of the most important knowledge about a building may still exist primarily in the heads of the people who operate it.

We have spent decades creating building information without necessarily creating building intelligence.

There is an important difference.

The physical building has to become the common reference point

Consider a fairly straightforward question: What exactly do I own?

For a commercial building, the answer should include its spaces, areas, equipment, systems, condition, configuration and the relationships between those components.

Now expand the question across a portfolio of 50, 100 or 500 properties.

Suddenly, answering it consistently becomes much harder.

Before sophisticated AI can reason about a portfolio, the industry needs a reliable digital representation of the physical assets themselves.

I think of this as digital ground truth.

It begins with accurately understanding the building as it exists today - not simply as it was designed years ago.

Modern reality-capture technologies such as LiDAR and 360-degree imagery can now document existing buildings quickly and with relatively little disruption.

That information can be converted into CAD, BIM and structured spatial information.

But digitization shouldn't end with creating another drawing or model. That should be the beginning.

A BIM model isn't necessarily a digital twin

The term digital twin has become widely used in commercial real estate. Sometimes too widely.

A detailed 3D model of a building can be extremely useful, but a model by itself isn't intelligent.

For an existing building, the larger opportunity is to connect its physical representation with the information required to understand and operate the asset.

A room isn't simply geometry. It has an area, use, tenant, finishes, equipment, maintenance history, energy consumption and potentially occupancy information.

A rooftop unit isn't simply an object in a model. It has a manufacturer, model number, installation date, service history, expected life, operating characteristics, sensor readings and replacement cost.

The relationships matter.

Once those relationships become structured and machine-readable, the digital representation of the building becomes considerably more valuable. It starts becoming an information architecture for the asset.

This is where AI gets interesting

Imagine an asset manager being able to ask:

Which HVAC units across our portfolio are approaching the end of their expected life?

Which buildings are consuming significantly more energy per square foot than comparable properties?

Where do our lease areas differ from recently verified areas?

Which assets have repeated maintenance problems?

What equipment should we replace during the next capital cycle?

What information are we missing about a building we're considering acquiring?

Today, answering those questions may require several people working across several systems.

Tomorrow, an AI agent should be able to answer many of them almost immediately.

But only if the information underneath that agent is trustworthy.

The goal shouldn't be another software silo

Commercial real estate already has substantial investments in property management, facility management, building automation, maintenance and enterprise software.

The answer isn't necessarily to replace those systems.

The larger opportunity may be to create an intelligence layer connecting them.

The physical building provides the spatial framework. Existing enterprise systems provide operational and financial information. Sensors provide real-time conditions. Historical records provide context. AI provides the ability to reason across all of it.

Instead of another application that people have to learn, the interface increasingly becomes a conversation with the building and its information.

From individual buildings to portfolio intelligence

This becomes considerably more powerful at portfolio scale.

One building can tell us something about itself.

A thousand consistently structured buildings can begin telling us something about buildings.

Patterns emerge. Equipment failures can be compared. Energy performance can be benchmarked. Maintenance strategies can be evaluated. Capital expenditures can be prioritized. Best practices discovered in one property can potentially be applied across hundreds of others.

That is where I believe some of the largest value will eventually be created.

The industry can move from building information to portfolio intelligence.

Building information should be treated as an asset

Building information has traditionally been created to complete a particular task.

Measure the building. Produce the drawings. Complete the BOMA calculation. Perform the condition assessment. Create the BIM model. Complete the acquisition. Then everyone moves on.

But what if the information created during those processes became part of a persistent digital asset that continued to improve throughout the life of the building?

A renovation updates it. An equipment replacement updates it. An acquisition adds another property. A condition assessment enriches it. Sensors make parts of it dynamic. AI makes it accessible.

Instead of repeatedly recreating building information, owners begin accumulating building intelligence.

That is a very different model.

AI may expose an uncomfortable truth

The AI revolution may reveal something the commercial real estate industry hasn't had to confront before.

Many organizations may not actually know as much about their physical assets as they think they do.

That wasn't necessarily a critical problem when information was being interpreted by people who could work around incomplete drawings, call the building operator or walk the property.

Machines are less forgiving.

AI requires structure, context and trustworthy information.

So before asking, “What can AI do for my buildings?”, owners may need to ask a more basic question:

Does AI actually understand my buildings?

Closing that intelligence gap may become one of the most important technology opportunities in commercial real estate.

Because in the future, the value of a commercial building may depend not only on the quality of the physical asset, but also on the quality of its digital representation.