AI in Commercial Real Estate: 74% of Firms Use AI Tools, $16.7B in Proptech Funding, and How Landlords Price Space

74% of commercial real estate firms now use at least one AI tool in core operations, up from just 39% in 2023, while proptech funding hit $16.7 billion in 2025 — a 67.9% jump from the year before. That's the short answer. The longer answer is that AI isn't just automating paperwork for landlords anymore; it's changing how space itself gets priced, in close to real time instead of once a quarter.

For most of commercial real estate's history, pricing space was a slow, relationship-driven process: a broker's gut sense of comps, a quarterly market report, and a lease renegotiation every three to ten years. That model is breaking down fastest in the two places AI touches most directly — data centers, where demand is exploding, and everywhere else, where landlords are using the same underlying tools to react to vacancy and tenant demand signals in weeks instead of quarters.

How AI in commercial real estate is changing how landlords price space

AI in commercial real estate now touches dynamic lease pricing, tenant demand forecasting, automated underwriting, and building operations, with 74% of firms using at least one AI tool in core workflows as of JLL's 2026 survey. That's nearly double the 39% adoption rate from just three years earlier, and it marks a shift from AI as an experimental add-on to a standard part of how deals get priced and underwritten.

The mechanics are straightforward: instead of a leasing team pulling comps manually once a quarter, AI-driven platforms ingest foot traffic, sublease listings, utility usage, and competitor asking rents continuously, then flag when a landlord's pricing has drifted from what the market will actually bear. That's a meaningfully faster feedback loop than the traditional broker-comp cycle, and it's showing up in how our Big Tech capex dashboard data center tenants get priced too — those leases increasingly get modeled with the same real-time-demand logic as office space, just against power and cooling capacity instead of square footage.

$16.7 billion in 2025 proptech funding — where the AI-native money actually went

Global proptech funding hit $16.7 billion in 2025, up 67.9% year over year, with capital increasingly concentrated around AI-native firms rather than the sector as a whole. AI-centered proptech companies grew roughly 42% annually in 2025, nearly double the ~24% growth rate for non-AI peers — the clearest sign yet that AI, not just a broader capital rebound, is driving the funding wave.

Four new proptech unicorns have emerged since mid-2024, and all four are AI-native: EliseAI ($2.2B valuation, AI leasing and resident communication), Bedrock Robotics ($1.75B, construction-site autonomy), Vantaca ($1.25B, AI-driven community association management), and Juniper Square ($1.1B, institutional real estate fund administration software). None of the four existed as unicorns before AI tooling became central to their product.

Smaller deals tell the same story: construction-site robotics startup Xpanner closed an $18 million Series B in May 2026, and New York-based Rebar raised a $14 million Series A to build AI-generated quoting tools for commercial HVAC suppliers. The money is flowing into narrow, workflow-specific AI tools rather than broad "AI for real estate" platforms — a pattern consistent with what we've seen across AI valuations more broadly, where vertical-specific products are commanding better multiples than horizontal ones.

Data center REITs: the one CRE sector AI is directly creating, not just improving

Data center REITs are the exception to "AI improves how you price existing space" — for this sector, AI is the demand itself. Hyperscaler AI capex is projected at roughly $700 billion in 2026 across Microsoft, Google, Amazon, and Meta combined, and every dollar of that assumes a landlord somewhere can deliver the physical shell, power, and cooling on schedule. Digital Realty closed the final round of its inaugural US hyperscale data center fund at $3.25 billion in total equity commitments on March 30, 2026, giving it dedicated capital to build exactly that.

That's a structurally different pricing dynamic than office or retail: data center landlords are pricing against power and cooling capacity constraints rather than square footage or foot traffic, and tenants (hyperscalers) are signing leases years in advance to lock in capacity rather than negotiating renewal terms on expiring space. It's the same underlying grid-capacity bottleneck we covered in our piece on grid modernization and practical climate infrastructure — the electrons, not the buildings, are now the scarce asset.

AI commercial real estate tools compared by function

The table below breaks down where AI is actually being deployed across commercial real estate, the scale of activity in each category, and what's driving it.

Figures blended from company funding disclosures, MarketScale, New Market Pitch, and Digital Realty press releases, 2025-2026. Valuations reflect most recently reported private-market figures.

Is the AI real estate market actually growing, or is this another hype cycle?

The AI-in-real-estate market was valued at roughly $303 billion in 2025 and is projected to reach $989 billion by 2029, a 34.4% compound annual growth rate, per current industry market sizing. That headline number is easy to write off as another AI market-size forecast, but the underlying adoption data backs it up: 74% AI tool usage across CRE firms and a 67.9% jump in proptech funding are both actual 2025-2026 figures, not projections.

The more useful signal for investors is where the AI-native growth premium shows up: proptech firms building AI-first grew about 42% annually in 2025 versus roughly 24% for peers still bolting AI onto legacy workflows. That gap is exactly the kind of "AI-native vs AI-bolted-on" divergence we've tracked across SaaS valuations more broadly — the companies rebuilding their product around AI from the start are outgrowing the ones treating it as a feature.

How dynamic pricing actually works for office and retail landlords

Away from data centers, the AI pricing shift looks less dramatic but is arguably more disruptive to how the industry has operated for decades. A traditional office landlord repriced a vacant floor by asking a broker to pull three to five comparable leases, adjusting for concessions, and presenting a range to an investment committee that met quarterly. That cycle could take four to six weeks from first signal to a revised asking rent, by which point the underlying market had often already moved.

AI-driven leasing platforms compress that cycle by ingesting sublease listings, foot-traffic data, utility consumption, and competitor asking rents on a rolling basis, then flagging pricing drift automatically instead of waiting for a scheduled review. That's the same underlying shift retail and hospitality already went through with dynamic pricing engines years ago — commercial real estate is simply the latest, slowest-moving asset class to adopt it, largely because leases are longer and deal sizes are larger, which historically made manual underwriting tolerable.

The tradeoff shows up in tenant negotiations too: sophisticated tenants and their brokers increasingly use the same class of AI tools to model a landlord's true vacancy cost and walk-away price before a renewal conversation starts, which narrows the information gap that used to favor landlords by default. That's pushing both sides toward faster, more data-grounded negotiations rather than the multi-month back-and-forth that used to be standard for anything above a small suite.

What this means for investors underwriting proptech and CRE exposure

For venture investors, the 42%-versus-24% growth gap between AI-native and non-AI proptech firms is the clearest underwriting signal in the sector right now: capital is rewarding companies that rebuilt their product around AI from the ground up over incumbents retrofitting a chatbot onto an existing workflow tool. That mirrors the pattern we've tracked across enterprise software broadly, where AI-native architecture — not just an AI feature checkbox — is what correlates with premium growth and premium multiples.

For real estate investors and family offices evaluating REIT exposure, the more actionable split is between data center REITs pricing against power capacity — a genuinely new, still-underpriced-risk asset class — and traditional office and retail REITs where AI is primarily a margin and speed improvement on an existing pricing model rather than a new demand driver. Both are real opportunities, but they carry different risk profiles: one is a bet on AI infrastructure demand continuing to outrun supply, the other is a bet on operational efficiency compounding over many lease cycles.