There is a phrase that has circulated through CRE executive suites for the past eighteen months with remarkable consistency. "We’re in the exploring phase." It shows up in earnings calls, at industry panels, in the careful language of operators who know AI matters but have not yet committed capital or restructured teams around it. The phrase has served as a socially acceptable way to acknowledge the shift without being held accountable for a timeline.
That phrase now has an expiration date.
In early March 2026, three independent signals converged to redraw the competitive map of AI in commercial real estate. Public Storage and Welltower, two REITs with a combined market capitalization approaching $199 billion, signed the first cross-sector AI licensing agreement in REIT history. Kroll launched REVS, an AI-enabled valuation platform built on more than $25 billion in annual institutional real estate appraisals across 15,000 commercial properties. And McKinsey published a report identifying agentic AI in property management and leasing as a $430 billion to $550 billion productivity opportunity, effectively putting a price tag on the gap between companies that have built AI infrastructure and those still exploring. None of these events happened in isolation. Together, they mark the moment AI crossed from operational experiment to balance sheet asset in commercial real estate.
This article is part of BestCRE’s ongoing coverage of AI-driven transformation across the commercial real estate industry. For a full view of how AI is reshaping every major property sector, explore our 20 CRE Sectors hub. Related analysis is available in our coverage of CRE Market Analytics & Data and CRE Underwriting & Deal Analysis.
The First REIT-to-REIT AI Deal Changes the Competitive Calculus
On March 1, 2026, Public Storage (NYSE: PSA) and Welltower (NYSE: WELL) announced a strategic data science partnership that has no precedent in the REIT sector. The structure is worth understanding precisely, because its implications extend far beyond the two companies involved.
Welltower built its data science platform in 2016, staffing it with a multidisciplinary team of Ph.D. computer scientists, engineers, statisticians, and mathematicians. That platform has powered more than $80 billion in capital allocation activity, compressing transaction timelines from the industry standard of five to nine months down to a matter of weeks using advanced mathematical models and high-performance computing. Welltower’s CEO Shankh Mitra has been explicit about the thesis: real estate has historically been a local, gut-feel industry, and the only way to truly scale it is through the data generated by the assets themselves.
Public Storage, for its part, has built differentiated operational data science capabilities as part of its broader transformation, including revenue management, customer behavior modeling, demand forecasting, and operating efficiency analytics. The company owns 3,533 facilities across 40 states representing approximately 258 million rentable square feet, generating a proprietary data set that competitors, third-party providers, and large language model interfaces simply cannot replicate.
The deal is bidirectional. Public Storage will license bespoke capital allocation models from Welltower, using supervised and unsupervised learning to focus on micro-markets with the greatest return and growth potential. Welltower will receive access to Public Storage’s operational analytics to drive performance improvements across the Welltower Business System. It is the first time Welltower has licensed a bespoke version of its platform to another operator.
What makes this structurally different from a typical technology partnership is how both companies framed it. This is not a vendor relationship. It is not a pilot program. The press release describes their respective AI capabilities as creating "a durable asymmetric information advantage," language that treats AI infrastructure as proprietary intellectual property with balance-sheet-level strategic value. Public Storage’s incoming CEO Tom Boyle and Welltower’s Mitra both used the word "durable" to describe the competitive moat these capabilities create.
For mid-market operators, the signal is actionable. The two largest companies in their respective sectors have concluded that sharing AI capabilities across asset classes generates more value than building in parallel silos. The question worth asking immediately is whether your organization has AI capabilities worth formalizing, and whether a partnership structure could accelerate sophistication faster than solo development. The companies building coalitions now will have infrastructure advantages in two to three years that solo builders will not be able to close.
Kroll REVS: When the Data Moat Becomes a Product
Two days after the Public Storage/Welltower announcement, Kroll launched the Real Estate Valuation Solution (REVS), an AI-enabled platform that automates commercial property valuations at institutional scale. The timing was coincidental. The thesis it validates is not.
Kroll valued more than $25 billion in institutional real estate across more than 15,000 commercial properties in the United States last year. That volume of transaction data, accumulated over nearly a century of operations, represents the kind of proprietary data advantage that cannot be replicated by a startup with a clever algorithm and a seed round. REVS is what happens when a firm sitting on that depth of data decides to productize it.
Ross Prindle, Managing Director and Global Head of Kroll’s Real Estate Advisory Group, was direct about the market forces driving the launch. Perpetual life funds, net asset value vehicles, and private wealth structures now demand more frequent, transparent, and data-driven valuations, creating operational pressure that many institutional investors are not equipped to handle with traditional appraisal workflows. REVS addresses this by combining portfolio insights, benchmarking tools, workflow automation, and appraisal management with metrics derived from Kroll’s comprehensive market indicators.
The platform’s competitive position illustrates a structural dynamic that will define AI adoption across every CRE function over the next five years: the hardest part of building a useful AI tool in real estate is not the AI. It is the data. Firms with decades of proprietary transaction data can build AI products that are difficult to compete with at a fundamental level. Platforms without that data advantage are building on inferior foundations, or they are licensing from the platforms that have it.
For appraisers and valuation professionals, the implication is immediate. The window to integrate AI-assisted workflows while maintaining competitive relevance is open now. It will not remain open indefinitely. The firms that adapt their processes before clients start asking why a valuation takes longer than a platform with AI-assisted automation will be positioned well. The firms that wait for the client question will find their answer insufficient.
McKinsey Names the Category: Agentic AI Gets a Price Tag
On March 4, McKinsey published a report on agentic AI in real estate that identified property management and leasing as the first verticals ripe for AI systems that execute multi-step workflows autonomously. These are not systems that answer questions or generate text. They take action within operational systems. The report estimated that automation applied to knowledge work, including agentic AI, could unlock roughly $430 billion to $550 billion in labor productivity across 48 countries.
The McKinsey report matters less for its specific findings (practitioners who have been building in this space already know where agentic workflows create value) and more for what its publication signals. When McKinsey names a category, board-level conversations in large CRE organizations accelerate. Budget approvals follow. The companies that have been building agentic capabilities for twelve to eighteen months will now see their slower competitors start to move.
The report describes a layered architecture for agentic AI deployment in real estate: an intelligence layer that ingests data and recognizes patterns, an action layer that executes work by integrating into property management and CRM systems, a control layer that manages permissions and audit trails, and a building-block layer of reusable agent routines. The specificity of the framework is itself a signal. This is not a conceptual paper about what AI might do someday. It is an implementation roadmap for organizations ready to commit engineering and operational resources.
For early adopters, the mainstream catching up is good news. It creates the ecosystem around them: more tools, more integrations, more talent familiar with the category. The head start does not disappear when the broader market arrives. It compounds.
The Supporting Signals Confirm the Pattern
The three primary events did not occur in a vacuum. Several supporting signals from the same period reinforce the same thesis: AI in CRE has crossed from discretionary experiment to structural competitive advantage.
Compass reports AI as a public markets narrative. Compass reported record 2025 revenue of $7 billion and disclosed that an enterprise-wide AI learning initiative launched just five months earlier had already identified approximately $20 million in potential annualized efficiencies, roughly 2% of Compass operating expenses. Anywhere Real Estate, which Compass recently acquired, is already processing approximately two-thirds of all brokerage documents through AI-driven automation, with its document assignment engine operating at 89% accuracy. The number that matters for CRE CFOs is not the $20 million itself. It is that AI ROI is now a public markets narrative. Companies that build the tracking infrastructure to quantify AI-driven savings will control how analysts value their technology investments.
Blackstone moves to democratize AI infrastructure exposure. Blackstone announced plans to launch a publicly traded acquisition company focused on buying leased AI data centers, approaching sovereign wealth funds for initial capital with the goal of eventually raising tens of billions from a broader investor base. Since taking QTS Data Centers private in 2021 for $10 billion, Blackstone has expanded QTS’s leased capacity fourteenfold. BREIT invested $5.8 billion in pre-leased data center developments in 2025, and expects substantially higher deployment in 2026. JLL estimates the data center sector will require up to $3 trillion in digital infrastructure investment by 2030. A publicly traded Blackstone data center vehicle would compete directly with Digital Realty and Equinix, giving institutional and retail investors direct exposure to the physical infrastructure powering the AI economy.
BXP creates a fourth category of property rights. BXP completed what it describes as the first formal transfer of digital property rights in a commercial real estate transaction. The $132 million December 2025 sale of a 409,000-square-foot office campus at 140 Kendrick Street in Needham, Massachusetts, to Lincoln Property Company and Cross Ocean Partners included a recorded blockchain transaction documenting control over how the property can be used in augmented and virtual reality environments. Neil Mandt, founder of Digital Rights Network (the platform through which BXP registered its entire portfolio), has described digital rights as a fourth category of property rights alongside air rights, mineral rights, and land rights. The revenue play involves AR advertising layered onto buildings visible through smartphones and smart glasses, a capability that could turn even warehouse assets along interstate corridors into monetizable digital canvases. The Digital Rights Network launched with more than $400 billion in registered real-world assets.
VTS hires a CPO from the hedge fund AI world. VTS, the dominant leasing and asset management platform in institutional CRE, appointed Adam Champy as Chief Product Officer. Champy comes from Point72, the global investment firm, where he served as Head of AI. Before that, he held product leadership roles at Google and Two Sigma. When someone with that caliber of AI-native product experience joins a CRE platform company, the product roadmap that follows will reflect a level of AI sophistication that CRE technology has not yet seen. The first major product announcement under Champy’s leadership will signal what he believes is the biggest unmet need in the market.
What the Convergence Actually Means for Capital Allocation
The significance of these events is not any single deal, product launch, or research report. It is the convergence: the fact that all of them happened in the same concentrated period, across different sectors, from different types of organizations, all pointing in the same direction.
When two $50-billion-plus REITs license AI capabilities to each other like intellectual property, the market is signaling that AI infrastructure has moved from cost center to asset. When a 100-year-old valuation firm productizes its data advantage into an AI platform, the market is signaling that institutional data is now a competitive moat with commercial value. When McKinsey publishes an implementation roadmap for agentic AI in real estate and attaches a half-trillion-dollar productivity figure, the market is signaling that the category has reached the point where management consultants can sell transformation services against it, which means budget cycles will follow.
For allocators and operators, the strategic question has shifted. It is no longer whether to invest in AI. It is where in the stack your competitive advantage sits. Do you have proprietary data worth formalizing as an asset? Do you have operational workflows where agentic automation could compress cycle times the way Welltower compressed transaction timelines from months to weeks? Do you have digital assets (physical properties with untapped AR, spatial computing, or IoT value) that remain unmonetized?
The companies answering those questions now are the ones who will not need to catch up in 2028. The ones still in the exploring phase will find that the frontier has moved without them.
The Partnership Question That Every Mid-Market Operator Needs to Answer
The Public Storage/Welltower deal will ripple through strategy conversations at mid-size operators for the next several months. Not because every operator needs to license AI from a megacap REIT, but because the deal establishes a template: AI capability can be shared, licensed, and co-developed across organizations and asset classes.
Most mid-market CRE firms do not have the resources to build a data science platform staffed by Ph.D. engineers. They should not try. The lesson here is not that everyone needs to build from scratch. It is that the build-versus-partner decision needs to be made deliberately, not deferred. Firms that identify what they are good at (whether that is operational analytics, tenant behavior modeling, or market-level pattern recognition) and formalize those capabilities as licensable assets will find willing partners. Firms that treat AI as a departmental initiative rather than a strategic capability will find themselves licensing from competitors or being acquired by them.
The exploring phase offered optionality. It allowed organizations to watch, learn, and avoid commitment. That optionality has a cost, and the cost is now visible. Welltower did not build its data science platform in 2026. It built it in 2016. The ten-year head start is now being monetized. The question for every other operator is how long they can afford to compound the disadvantage before the gap becomes structural.
Where CRE AI Goes From Here
Early March 2026 will likely be remembered as the moment CRE’s AI adoption curve bent. Not because any single event was unprecedented in isolation, but because the clustering of signals made the direction unmistakable.
Watch for three developments in the next sixty to ninety days. First, whether other REITs follow the Public Storage/Welltower template and announce cross-sector AI partnerships. If two more surface by Q2, the coalition model becomes the industry standard. Second, whether Kroll REVS triggers competitive responses from CBRE, JLL, or Cushman & Wakefield in the valuation automation space. The firms with proprietary appraisal data will either build or lose market share to those who have. Third, whether VTS’s first major product release under Adam Champy introduces agentic capabilities. The CRE platform that embeds autonomous workflow execution into leasing and asset management will reset expectations for every other vendor in the category.
The operational advantage in commercial real estate is no longer defined by who has the most properties or the best locations. It is increasingly defined by who can process information faster, underwrite more accurately, and operate with less friction. That race is underway. The signals from March 2026 made it obvious who is running it, and who is still standing at the starting line.
BestCRE.com is the leading platform for commercial real estate AI intelligence, market analysis, and investment strategy. We cover the tools, transactions, and trends shaping the future of CRE across 20 industry sectors. For AI tool reviews, institutional market analysis, and data-driven perspectives on where capital is flowing, explore our complete coverage.
Frequently Asked Questions
What is the Public Storage and Welltower AI partnership?
Announced on March 1, 2026, the Public Storage/Welltower partnership is the first cross-sector AI licensing agreement between two major REITs. Public Storage will license bespoke capital allocation models built by Welltower’s data science platform, which has powered more than $80 billion in capital allocation activity since 2016. In exchange, Welltower gains access to Public Storage’s operational analytics capabilities, including revenue management, demand forecasting, and customer behavior modeling. The combined market capitalization of the two companies approaches $199 billion, making this the largest AI-focused partnership in REIT history.
How does Kroll REVS change commercial real estate valuation?
Kroll REVS is an AI-enabled platform that automates commercial property valuations at institutional scale. It combines Kroll’s portfolio insights, benchmarking tools, and workflow automation with metrics derived from its comprehensive market data set, informed by more than $25 billion in institutional real estate valuations across 15,000 commercial properties annually. The platform addresses growing demand from perpetual life funds, NAV vehicles, and private wealth structures for more frequent, transparent, and data-driven valuations, reducing cycle time while maintaining audit-ready documentation and regulatory compliance.
What is agentic AI and why does it matter for CRE?
Agentic AI refers to artificial intelligence systems that can execute multi-step workflows autonomously within operational systems, moving beyond the question-and-answer capabilities of generative AI to take direct action: creating work orders, scheduling vendors, updating records, and routing approvals. McKinsey’s March 2026 report identified property management and leasing as the first CRE verticals ripe for agentic deployment, estimating that automation including AI applied to knowledge work could unlock $430 billion to $550 billion in labor productivity across 48 countries. For CRE operators, agentic AI represents the shift from tools that inform decisions to systems that execute them.
What are digital property rights and how did BXP create the category?
Digital property rights give building owners control over how their properties are represented and used in digital and virtual environments, including augmented reality overlays, spatial computing experiences, and location-based advertising. BXP completed the first formal transfer of these rights in a $132 million December 2025 sale of a Needham, Massachusetts office campus. The rights were documented via blockchain on the Digital Rights Network platform, which launched with more than $400 billion in registered real-world assets. The revenue potential lies in AR advertising, where brands can place digital billboards on buildings visible through smartphones and smart glasses without physical signage.
How should mid-market CRE operators respond to the AI acceleration?
The key lesson from early March 2026 is that the build-versus-partner decision can no longer be deferred. Welltower’s data science platform was built in 2016, and the ten-year head start is now being monetized through licensing. Mid-market operators should identify their specific data and analytics strengths, evaluate whether those capabilities can be formalized as licensable assets, and determine whether partnership structures could accelerate AI sophistication faster than solo development. Compass’s disclosure that a five-month AI initiative already identified $20 million in annualized efficiencies (2% of operating expenses) provides a benchmark for what early-stage AI adoption can deliver at scale.
Related articles:
Best CRE Data Centers: Why Power Is the New Location
Best CRE Healthcare: Why AI Is the New Demographic
Best CRE Office Market: Bifurcation, Not Recovery
