CRE Underwriting Meets AI: Navigating Capital Allocation and Data Gaps

By RadCRE Research · · Market Updates

While 83% of CRE investors deploy AI, JLL research finds capital allocation lags operational gains as underwriting faces tighter scrutiny.

Capital Deployment Versus Operational Realities in CRE AI

Institutional adoption of artificial intelligence across commercial real estate has reached critical mass, yet capital allocation strategies are encountering an unexpected performance gap. According to research released by JLL, 83% of real estate investors have advanced at least one AI initiative beyond the initial testing phase [1]. A significant portion of these market participants—49%—aim for AI to drive outperformance and strengthen capital deployment capabilities [1].

However, the measurable returns from these tools are not landing where top-level investment committees initially expected. JLL's data shows that asset and property management functions are generating the strongest measurable gains, delivering a 66% increase in operational efficiency and a 57% improvement in decision quality [1]. While underwriting and acquisition teams seek to deploy machine intelligence to outmaneuver competitors on transaction volume and pricing, operational workflows are currently realizing the primary benefits of automation [1].

Underwriting Pressures Mount Amid Refinancing and Stricter Terms

The gap between underwriting ambitions and execution comes at a challenging juncture for commercial real estate balance sheets. A looming multifamily refinancing crunch is raising the stakes for AI deployment across capital markets [3]. Simultaneously, regional property markets are confronting the headwinds of stricter underwriting standards from lenders [4]. In regions such as California's Central Valley, the Fresno-Clovis multifamily sector faces the dual friction of tighter underwriting criteria and moderating rent growth as operating costs rise [4].

In this environment of disciplined lending and elevated cost structures, legacy underwriting models struggle with compressed margin profiles. While investment teams seek algorithmic outperformance to identify risk-adjusted returns, flawed underlying inputs frequently bottleneck AI-driven underwriting models before actionable insights emerge.

The Agentic Workflow and Data Readiness Divide

JLL's findings underscore that access to standard foundational AI models is now merely an industry baseline rather than a competitive edge [1]. The true performance divergence lies in data infrastructure and autonomous workflow design. Only 15% of institutional real estate investors have developed custom agentic workflows across all their AI deployment areas [1]. Those firms that have engineered custom agentic workflows are more than twice as likely to report broad institutional benefits across their portfolios [1].

Data readiness remains the definitive filter between successful execution and stranded technology spend. Among investors whose data is not AI-ready, 59% struggle to realize tangible benefits, compared to only 21% of firms operating with fully AI-ready data architectures [1]. Without normalized historical performance figures, clean rent rolls, and synchronized property metrics, underwriting engines face persistent data friction.

RadCRE Perspective

"The fact that 83% of investors have adopted AI proves technology access is no longer a differentiator in commercial real estate," says Majid Radaei, Founder and Principal Broker of RadCRE. "The real divide is between firms with clean, AI-ready data infrastructure and those feeding unstructured noise into generic algorithms. With stricter underwriting standards taking hold and multifamily facing a significant refinancing test, automated agentic underwriting workflows that bridge property-level data directly into capital stack analysis will separate deal-closers from the rest of the market."

As debt maturities force complex recapitalizations, real estate investment banking teams must pivot from passive software experimentation to structured, agentic underwriting pipelines that directly connect property-level operating efficiencies to rigorous capital stack execution.

Tags: commercial real estate AI underwriting, institutional CRE capital deployment, multifamily debt refinancing, CRE agentic workflows

Sources (published in the past 7 days):