AI’s Growing Impact on Commercial Real Estate

The industry is still in the early stages of realizing AI’s potential while also wrestling with its downsides.

The world is captivated by the potential of artificial intelligence. Both companies and individuals are experimenting with accessible tools like ChatGPT, a chatbot developed by OpenAI that is trained to provide humanlike responses to user prompts. CRE companies are using AI to improve efficiency in everything from lease management to building operations. A 2023 report from Fortune Business Insights predicts that the proptech market will expand from $34 billion in 2023 to $90 billion by 2032.

However, the rise of AI has also exacerbated age-old concerns over technology’s displacement of workers and questions about the amount of power it takes to generate even the simplest of commands. Despite its corporate sustainability initiatives, Google’s greenhouse gas emissions increased 48% over the last five years due to the integration of AI into its products. A Washington Post report in September calculated that one 100-word email generated by ChatGPT requires the equivalent of one bottle of water to cool the servers on which these bots run.

But most of the analysis surrounding AI’s impact on the commercial real estate industry — whether positive or negative — is focused on its potential. In these early stages of the AI revolution, it is difficult to decipher exactly how it is being applied or its overall effect. Regardless, conversations with leaders in the industry suggest that AI’s influence on commercial real estate is beginning to take root and spread.

Occupier Solutions

Chris Zlocki is head of client experience and executive vice president, occupier solutions, at Colliers. This past summer, he led a report, AI in CRE: The Now, the Next and the Possible, examining AI’s impact on key areas of occupier services, including lease administration, facility management, workplace advisory, portfolio strategy, location intelligence, transaction management, project management and supply chain solutions.

“What used to take a lease administration team five to seven days now takes minutes,” Zlocki said in relation to AI’s ability to process huge quantities of documents.

Lease administration is where much of the advancement in AI can already be seen. Colliers360, the company’s integrated analytics platform, has incorporated AI into its suite of applications. Zlocki said AI’s immediate impact involved reducing the redundancy and increasing the speed of task completion. Additionally, the platform’s ability to comprehend and process data has improved dramatically.

“We have a client who was the victim of a cyberattack,” Zlocki said. “Their entire system was disabled, and they were at risk of falling behind on lease and vendor expenditures. We had to get notices out to landlords to let them know what had happened and why the client would be delayed in making their regular payments.”

Colliers used AI optical character recognition software, which converts image-based text into machine-readable format, to analyze 40 to 50 screenshots of printed rent rolls and lease documents. The information pulled from these screenshots included payment amounts and to whom payment was due. The client, a national energy company, didn’t miss a payment across its entire portfolio of leases.

“AI was able to instantly do what would have taken two to three people a week to accomplish,” Zlocki said. “And its ability to predict certain outcomes, and make recommendations based on those predictive outcomes, is changing the way clients are addressing their real estate decisions.”

Colliers’ Portfolio AI tool, developed by the company’s Portfolio Strategies and Technology Services divisions, looks at a company’s real estate obligations from a portfolio-wide perspective and makes an assessment based off the data it receives. It searches for opportunities to optimize a client’s facility portfolio, churning out recommendations to improve overall performance and operational efficiency.

In one case, Zlocki said, Portfolio AI’s assessment led to a recommendation that a client change up to 40% of its office portfolio planning decisions to reach optimal performance. Portfolio AI assessed the client’s real estate footprint using different financial, real estate and operational scenarios to determine cost-saving strategies for both individual leases and across the entire portfolio.

AI is also expected to have a significant impact on warehouse and third-party logistics. A January 2024 report from accounting and consulting firm EY asserts that 40% of supply chain organizations are investing in generative AI. The report identifies key areas like demand forecasting, risk management, and logistic network design and route optimization as sectors where AI will bring value to supply chains. Backed with this abundance of cognitive data, warehouse users will be able to make more informed space-related decisions for both their short- and long-term needs.

Practicing Quality Assurance

These early examples of the technology’s value have raised expectations for AI in the commercial real estate industry, Zlocki said. But to reach the full potential of what AI can contribute, he added, industry stakeholders first need to figure out how they are going to collect and share the data that will inform AI. That means encouraging transparency among firms and ensuring the data is accurate and up to date. It also means instituting some type of data governance and creating parameters around how the data will be used.

Zlocki said a new profession, AI ethicist, is emerging to set guidelines around the uses and purpose of AI. As AI’s technical capabilities evolve, becoming more intuitive and intelligent, how the technology is used will need to be reined in.

President Biden issued an executive order in 2023 outlining guidelines for the responsible collection of data used by tech companies in training AI models. The order serves as a set of principles for ethical AI application, especially in employee surveillance, inherent bias and the use of privacy-enhancing technologies in collecting data to train AI machines. Companies pledging their support of the guidelines include Apple, Amazon, Google, OpenAI and Microsoft.

Building Design and Construction

Michael Villegas of IA Interior Architects is using AI to produce high-quality early visuals of design projects for his office and industrial clients. Villegas leads the international design firm’s AIRIA team, a research group that focuses on AI’s potential to enhance design and project processes. Advanced visuals aren’t uncommon, he said, but they typically come later in the project timeline due to the time and cost it takes to produce them. AI shortens the time frame by 30%, according to Villegas, and the ability to provide a 3D rendering, based on conceptual planning alone, streamlines clients’ decision-making and accelerates project timelines.

“Designers can now use AI to generate hundreds of potential solutions in minutes,” Villegas said. Five years ago, he added, the same process would have taken weeks. “The process was so linear back then. Now it’s much more iterative and flexible.”

Villegas also emphasized AI’s ability to democratize projects by collecting prompts and inputs from multiple project stakeholders to inform the design process. He listed programs like ChatGPT, Midjourney (text-to-image AI) and various generative AI plug-ins for traditional software like Autodesk Revit as tools already in use by architects at IA Interior Architects and elsewhere.

AI is displacing some specialties traditionally included in the design and development process. But Villegas said what AI is actually replacing is the repetitive tasks that tend to bog designers down, opening opportunities for architects and designers to put their creativity to better use.

His advice is to approach AI with clear problems for it to solve or goals for it to achieve. “AI complements human expertise, it doesn’t replace it,” Villegas said. “The best results come from a partnership between AI-driven insights and human intuition, innovation and problem-solving capabilities.”

Supporting Building Operations

Building management systems are already proving AI’s sustainability benefits. The Colliers AI in CRE report references numerous owners implementing AI into their management platforms to harness sensor data, understand usage patterns and predict instances of energy wastage.

“Given the prominence of the green revolution, carbon neutrality, and the ongoing emphasis on energy, it will be imperative for our [Facility Management] Advisory team to draw insights from these evolving AI applications to provide advice and guidance to clients,” the Colliers report reads.

Matthew McAuley, global property sectors research senior director for JLL, said that with the use of AI, building management systems can adjust their heating and cooling presets based on current climate and weather forecasts, reducing waste and embodied carbon.

“JLL’s Hank platform is one of these software systems that can be integrated into building management systems,” he said. “Hank leverages AI to adjust daily operations and settings to maximize performance and reduce energy use. This can lead to reductions in energy costs of 20% or more.”

Nome Capital Partners, a Bay Area-based real estate firm primarily focused on investments in office properties, used JLL’s Hank in a highly leased building to reduce HVAC energy consumption by 45%, achieve 60% less deviation from tenant setpoint, and increase its circulation of fresh air to double the minimum amount required — improving indoor air quality and the health of the building for occupants overall.

Zlocki added that facilities management is relatively low-hanging fruit for AI due to the growth in Internet of Things (IoT) technology across the commercial real estate industry. The combination of AI and IoT leads to a building being able to tell an owner or manager which part of the structure needs what service and how much.

“We’re even getting to the point with AI that a building can start to predict maintenance and repair needs,” Zlocki said. “The building can perform tests on itself, becoming more prescriptive rather than reactive.”

Impact on Office Space Demand

McAuley doesn’t expect that AI will cause lease rates to rise or drop or that it will affect other fundamentals like term length. However, all eyes are watching how AI’s expanding prevalence will affect job numbers, which heavily influence building occupancy rates. People need space to work — whether office space or elsewhere — and strong employment means more expansion and more space leased. With AI? Not so much.

McAuley puts it bluntly: “AI will have an effect on most jobs. For example, the International Monetary Fund estimates that 40% of all jobs globally, and 60% in advanced economies, involve tasks that can be augmented or automated with AI.”

However, McAuley said it will take some time before any real effects are felt. “Most AI used to date has been experimental, entry level or specific to certain occupations,” he said. “The eventual impact on employment in specific industries is still highly uncertain. While specific tasks within jobs can be augmented or automated, it’s unlikely that many jobs can be fully automated with existing AI capabilities.”

Echoing Villegas’ sentiment about the ways that AI can enhance the work of designers, McAuley thinks some employees — especially those with a high degree of repeatable tasks in their descriptions — might be more productive at their jobs with assistance from AI.  

“Technology has historically created more jobs than it removes,” he said. “Sixty percent of current employment is in occupations that didn’t exist 80 years ago. This process may lead to more job moves, but the net effect on employment will remain unclear for some time.”

McAuley noted a fairly obvious exception where AI’s impact is already evident in the real estate market: the tech sector. And the geographic area most affected? The Bay Area.

In San Francisco, the traditional epicenter of tech activity, AI companies have leased 1.7 million square feet of office space since 2022, bringing the total AI office footprint in the city to 4.3 million square feet. The largest lease transaction in the fourth quarter of 2023 was OpenAI, the developer of ChatGPT, which leased 486,000 square feet of office space in the city’s Mission Bay submarket, adding to the 140,000 square feet the company has across multiple leases elsewhere in the metro area. Other deals include AI startup Anthropic’s lease of 230,000 square feet in the South Financial District and autonomous vehicle company Zoox’s 219,000-square-foot industrial lease in the East Bay. In nearby San Jose, Couchbase, an AI-powered cloud database, signed a lease in August 2024 for 24,000 square feet.

McAuley suggested these recent AI expansions are promising indications of what else may come. “In the U.S., the footprint of AI companies has more than doubled since 2020,” he said. “AI companies are also active beyond the office sector, with robotics and autonomous vehicles companies leasing industrial space or AI drug discovery companies taking lab space.”

Blanket optimism isn’t advised, however, especially given that layoffs at tech companies have been the norm for the past couple years. According to Crunchbase’s Tech Layoffs Tracker, 191,000 tech workers were laid off in 2023, and at least 94,733 jobs had been lost in the tech sector in 2024 through Nov. 15. This broader tech industry trend, accompanied by uncertainties related to the ongoing remote versus hybrid versus in-person work debate, doesn’t necessarily signal a return to pre-pandemic office occupancies.

Potential for Improved Investment Performance

While AI is showing up on some owners’ rent rolls, it’s also making its way into their spreadsheets.

“AI tools are increasingly being integrated into investment decision-making,” McAuley said. “Key use cases include market and asset-level analytics and forecasting, risk assessment, automated or enhanced valuations, and asset filtering and identification.”

He points to JLL’s Capital Markets Quants platform as an example. The platform “analyzes data from over 1.25 million properties transacted globally in the past 20 years to predict shifts and opportunities in the real estate market, giving advisers more accuracy and intelligence in optimizing their clients’ portfolio strategies.”

Again, AI is doing the job that real estate investors and analysts have always been doing, but at a speed that allows owners to make complicated, high-volume investment decisions early in the transaction process and armed with more information. Its predictive capabilities can analyze the market for opportunities and risk factors, giving investors a heightened level of certainty before deciding whether to proceed with a transaction. Conceivably, that additional certainty should improve investment performance.

But as reliance on AI grows, it creates its own risks. McAuley identified three primary areas of concern AI presents for the real estate market: data, privacy and security risks; regulatory and compliance risks; and operational risks. He said most countries will have AI-related legislation by 2030, driving transparency in the training and use of AI.

This goes back to an observation both Zlocki and Villegas made: For AI to be a successful tool for the real estate industry, there must be transparency among its stakeholders. Skepticism of AI can be countered through proper training and accessibility to the data AI is consuming and producing.

Early Promise and Ongoing Questions

The AI revolution is just beginning to gain momentum. Much of the literature emphasizes its potential rather than its current effects. But it is already helping companies process data at speeds not seen before. That ability to instantaneously analyze large quantities of information and present scenarios and recommendations for users is remarkably useful to the real estate industry.

Serious questions remain, however, about the environmental and societal implications of broad acceptance of AI. A regulatory framework and a definition of ethical use of AI are needed to ensure the technology is a net positive rather than a detriment.

Will McDonald is a freelance writer in Brooklyn with prior professional experience as a commercial real estate broker.

AI’s Effect on Data Centers JLL anticipates global co-location and hyperscale data center inventory will grow at an annual rate of 20% over the next three years. That means by 2027, data center demand will more than double from current capacity. A report from Goldman Sachs projects that by 2030, demand for data center power will increase 160% due to wider implementation of AI. Major investments in data center infrastructure will also be required to meet the rising AI demand: $50 billion in new power generation in the United States, and nearly $1 trillion to ready Europe for the oncoming AI surge.

Artistic Inspiration This AI-generated rendering was produced with guidance and processing from IA Interior Architects in collaboration with feedback from project stakeholders. Rendering courtesy of IA Interior Architects IA Interior Architects worked with a community organization looking to design and install a public art display in Miami’s Little River neighborhood, an up-and-coming area that wanted to commemorate its creative roots with public art reflecting the community. The process began with a vision session to gather the client’s aesthetic preferences, project goals and unique understanding of the neighborhood through audio, text and images. IA Interior Architects fed the feedback from that vision session into different AI systems to analyze patterns to unlock a distinct creative direction for the project. This allowed the client to effectively design the mural themselves in real time and save money on third-party design consultants and illustrators. “The AI outputs provided the source material for a human-crafted collage, which we presented to the client,” said Michael Villegas of IA Interior Architects. “This collage reflected both the client’s vision and local values, ultimately leading to an expanded scope that benefited not just the project but the community.”

• Magazine Article
• Business/Trends
• Facility Management
• Technology
• Design
• Artificial Intelligence (AI)