AI Reaches Its Execution Phase in Commercial Real Estate

Artificial intelligence is no longer an emerging concept in commercial real estate. It is being tested, piloted and debated across investment committees, operating teams and boardrooms throughout the industry. Yet despite growing optimism and accelerating investment, most firms remain stuck in early-stage deployment.

That disconnect is the central finding of the 2025 State of AI Adoption in Real Estate Survey, conducted by Keyway in partnership with the Appraisal. The data paints a clear picture of an industry that believes deeply in AI’s promise, but continues to struggle with execution. The challenge facing real estate today is no longer whether AI belongs in the enterprise, but how to move it from experimentation into durable, scaled deployment.

The survey reveals a growing gap between intent and impact. While nearly half of firms report running AI pilots, only a small fraction have deployed AI across the enterprise. Even fewer say their data infrastructure is fully prepared to support AI at scale. This imbalance is not simply a technical shortcoming. It reflects a strategic failure to align data, systems and decision-making processes around AI as a core operating capability rather than a series of isolated experiments.

As a result, firms are committing real capital to pilots that never progress beyond proof of concept. These efforts accumulate cost without compounding value, while competitors that solve execution gain speed, efficiency and decision-making leverage. AI adoption has quietly shifted from a question of willingness to one of readiness. Data infrastructure, not organizational enthusiasm, has become the primary constraint.

Trust represents a second, equally important barrier. While AI is increasingly accepted for efficiency-driven tasks, its use in high-stakes financial decisions remains limited. A significant portion of respondents report that investment committees distrust AI-generated analysis, and only a minority express confidence in using AI for underwriting. Concerns consistently center on explainability, integration with legacy systems, data quality and the risk of unreliable outputs.

This skepticism helps explain why AI adoption has plateaued at the operational layer. For AI to move upstream into valuation, underwriting and capital allocation, it must produce outputs that are explainable, auditable and grounded in traceable data. Sophistication alone is insufficient. Trust, not technology, is the gating factor.

Despite persistent narratives around automation-driven job displacement, the data suggests a far more pragmatic reality. Most firms report using AI to improve efficiency rather than reduce headcount, with relatively few anticipating workforce reductions tied to adoption. In practice, AI is being deployed to eliminate manual work, accelerate analysis and increase operational leverage. The prevailing model is augmentation rather than replacement, with AI functioning as an invisible force multiplier behind faster, more effective teams.

Perhaps the strongest signal of AI’s permanence is financial. A majority of surveyed firms plan to meaningfully increase AI spending over the next two years, and more than half expect to purchase new AI software within the next 12 months. Notably, these investments are being funded primarily through reduced outsourcing and administrative costs, not workforce cuts. AI has crossed an important threshold. It is no longer a discretionary experiment but a core technology line item embedded in operating budgets.

Adoption patterns vary meaningfully by asset class. Student housing currently leads in enterprise deployment, while office shows high experimentation though limited conversion to scale. Multifamily, despite being the largest and most operationally complex asset class, exhibits the lowest level of enterprise-wide AI adoption. That gap represents one of the sector’s most significant opportunities. Given multifamily’s scale, data intensity and margin sensitivity, it is arguably the vertical best positioned to benefit from AI-driven underwriting, comps analysis and asset management, once the necessary data infrastructure is in place.

Across every segment, one constraint dominates. Fragmented systems and inconsistent data continue to limit AI’s effectiveness, with most firms acknowledging significant gaps in their data foundations. The implication is straightforward. The winners in AI-enabled real estate will not be those who deploy the most tools, but those who standardize data and embed intelligence directly into core workflows.

AI adoption in real estate is inevitable. Scalable adoption is not. The next wave of competitive advantage will belong to firms that pair clean, structured data with explainable AI and integrate it directly into underwriting, asset management and risk analysis. The experimentation era is ending. The execution and monetization era has begun.

Over the next 24 months, firms that invest in data readiness, prioritize trust and operationalize AI within core workflows will define the industry’s next competitive hierarchy. Those that do not risk remaining permanently stuck in pilot mode.

Ashkán Zandieh is the founder and managing director at the Center for Real Estate Technology & Innovation (CRETI). Nima Wedlake is the managing director at V1 Ventures and a venture member at CRETI.