Enterprise AI Is Turning Proptech Into an Infrastructure Market
The signal in commercial real estate is not that proptech startups are being pushed aside. It is that AI has moved from the innovation budget to the operating model. As Commercial Observer reported, major capital groups and landlords are now backing enterprise AI deployments at a scale that would have looked unlikely in real estate only a few years ago.
The numbers matter. Proptech venture investment reached $16.7 billion in 2025, up 68 percent year over year, according to data cited from the Center for Real Estate Technology and Innovation. AI-native proptech companies captured $4.5 billion of that capital. This is not a retreat from the category. It is a rotation within it, away from conventional SaaS tools and toward systems that can interpret documents, connect fragmented data, automate workflows and improve portfolio decisions.
The largest deals sharpen the pattern. Anthropic’s $1.5 billion Ode venture, backed by financial players including Blackstone and Goldman Sachs, is designed to help firms customize and support AI deployments. OpenAI’s $10 billion deployment venture with TPG, Brookfield, Bain Capital, Advent, SoftBank and Dragoneer is even more explicit about real estate portfolio companies. These are not simple software purchases. They are attempts to build AI capability into the machinery of ownership, underwriting, asset management and operations.
For KG Data readers, the key question is where value will concentrate. Real estate’s biggest technical problem is still data fragmentation. Lease data, valuation assumptions, maintenance records, construction updates, local market definitions and transaction histories often live in different formats and systems. AI can summarize and classify information, but weak data architecture limits model performance. The firms that solve identity, normalization, permissions and auditability may create more durable value than those offering narrow workflow automation.
AI does not eliminate proptech’s fragmentation problem. It makes the cost of fragmented data more visible.
This explains why incumbents still matter. CoStar, Yardi, RealPage, Procore and AppFolio collectively generated about $9.2 billion in fiscal 2025 revenue, according to the article. Their advantage is not only software adoption. It is embedded data, customer workflow history and distribution. AI may reduce the defensive value of old interfaces, but it increases the strategic value of proprietary datasets and trusted system-of-record positions.
Large owners building internal AI tools should therefore be read as market maturity, not startup displacement. Blackstone and Brookfield are unlikely to become horizontal software vendors. Their incentive is portfolio intelligence: faster underwriting, better risk detection, more consistent reporting and operational efficiencies across complex holdings. That creates demand for external vendors, model specialists and data infrastructure providers that can plug into enterprise environments.
The more interesting pressure falls on legacy SaaS companies that mainly wrap existing workflows without owning unique data or delivering measurable intelligence. In an AI-first market, buyers will ask harder questions: Does this tool reduce decision latency? Does it improve forecast accuracy? Does it integrate with existing asset, lease and financial data? Can its outputs be audited? The next proptech cycle will be judged less by dashboard polish and more by model reliability, data depth and operational impact.
What should investors, operators and builders track next? Watch AI-native funding share, enterprise deployment renewals, integration partnerships with system-of-record platforms and evidence of productivity gains in underwriting, construction and property operations. The headline deals are large, but the more important signal is structural: commercial real estate is beginning to treat technology as core infrastructure rather than optional enhancement.
Source: Commercial Observer
