Key Takeaways
- • Dealpath's 2026 State of AI in CRE Investing survey of 103 investment and technology professionals at U.S. and Canadian institutional real estate firms (managing $500 million to more than $40 billion) found that 97 percent say AI is integrated into their investment process, but gains remain modest and uneven.
- • Verification is nearly universal: 94 percent of respondents check AI outputs most of the time or always, and only 51 percent say AI saves time once verification is counted, while 41 percent say AI-assisted work now takes longer than manual analysis.
- • Fragmented deal data is the most-cited reason AI falls short (43 percent), ahead of hallucinated outputs (36 percent), and 90 percent say data quality or fragmentation has limited AI's impact at their firm despite 83 percent rating their data infrastructure as ready.
- • Trust declines as tasks move toward valuation: 55 percent highly trust AI to summarize diligence documents, but only 35 percent highly trust it to score and rank opportunities.
- • Dealpath argues firms should pay a one-time 'data toll' to structure deal data rather than an ongoing 'verification tax', and that proprietary deal data and recorded decisions will drive future AI returns.
Summary
Institutional real estate investors have spent several years approving AI budgets and embedding AI into sourcing, screening, underwriting and asset management. According to a new report from Dealpath, titled The 2026 State of AI in CRE Investing: Adoption Without Impact, the question has shifted from adoption to whether AI is changing how firms invest. Dealpath commissioned independent research firm Censuswide to survey 103 investment and technology professionals, from analysts to the C-suite, at institutional real estate firms in the United States and Canada. Fieldwork was completed in the second quarter of 2026. The report's headline conclusion is that the gains reported so far have been modest and uneven.
Adoption has outpaced measurable outcomes. Nearly all respondents (97 percent) said AI is integrated into their firm's investment process, and 52 percent described it as fully integrated. The outcome most often cited as improved was reduced time on manual and repetitive tasks, named by 36 percent, followed by faster deal cycles at 35 percent. Only 19 percent cited more precise investment decisions. Dealpath notes that these top gains accelerate individual tasks, such as extracting figures from documents or summarizing diligence materials, within otherwise unchanged workflows.
Verification absorbs much of the time saved. Ninety-four percent of respondents said they check AI outputs most of the time or always. Dealpath calls the cost of confirming AI outputs are accurate enough to act on the 'verification tax,' a cost that rarely appears as a line item. Only 51 percent said AI saves time once verification is counted, while 41 percent said AI-involved work takes longer than building the same analysis manually. The burden varies by function: 51 percent of respondents in credit said AI makes their work slower, while 59 percent of those in dispositions said it saves time.
The survey also points to a confidence gap around data. While 83 percent of respondents rated their firm's data infrastructure as mostly or fully ready for AI, 90 percent said data quality or fragmentation has limited AI's impact. Incomplete, inconsistent or fragmented data was the most-cited cause of AI shortfalls at 43 percent. Investment professionals were more likely than technology and data leaders to say data has limited AI (77 percent versus 56 percent). Dealpath's explanation is that AI requires a higher bar than professionals working around data gaps, since a model inherits gaps and fills them with guesses rather than flagging them.
Trust varies by task. Informational tasks earned the most trust, with 55 percent highly trusting AI to summarize diligence documents and 53 percent to screen opportunities for baseline fit. Judgment tasks closest to valuation earned the least, with 35 percent highly trusting AI to score and rank opportunities and 36 percent to recommend comparable properties. Underwriting had the lowest adoption of any lifecycle stage at 48 percent, yet 23 percent named it as the stage where AI has delivered the most measurable value, second only to portfolio management at 31 percent.
Dealpath frames the outlook as a choice between two costs: the recurring verification tax on outputs built on fragmented data, and a one-time data toll to structure, connect and govern deal data. The report also cites 71 percent of respondents using general-purpose large language models directly, arguing that a firm's competitive edge will therefore lie in proprietary deal flow and recorded investment decisions. Dealpath's 2025 State of AI Readiness survey of 100 institutional investors, cited for comparison, found 61 percent expected AI to deliver faster deals; in 2026 only 35 percent named deal speed among the most-improved outcomes. The report concludes that the return on AI in 2027 will depend more on the context firms provide to models than on which model they use.
Adoption has outpaced measurable outcomes. Nearly all respondents (97 percent) said AI is integrated into their firm's investment process, and 52 percent described it as fully integrated. The outcome most often cited as improved was reduced time on manual and repetitive tasks, named by 36 percent, followed by faster deal cycles at 35 percent. Only 19 percent cited more precise investment decisions. Dealpath notes that these top gains accelerate individual tasks, such as extracting figures from documents or summarizing diligence materials, within otherwise unchanged workflows.
Verification absorbs much of the time saved. Ninety-four percent of respondents said they check AI outputs most of the time or always. Dealpath calls the cost of confirming AI outputs are accurate enough to act on the 'verification tax,' a cost that rarely appears as a line item. Only 51 percent said AI saves time once verification is counted, while 41 percent said AI-involved work takes longer than building the same analysis manually. The burden varies by function: 51 percent of respondents in credit said AI makes their work slower, while 59 percent of those in dispositions said it saves time.
The survey also points to a confidence gap around data. While 83 percent of respondents rated their firm's data infrastructure as mostly or fully ready for AI, 90 percent said data quality or fragmentation has limited AI's impact. Incomplete, inconsistent or fragmented data was the most-cited cause of AI shortfalls at 43 percent. Investment professionals were more likely than technology and data leaders to say data has limited AI (77 percent versus 56 percent). Dealpath's explanation is that AI requires a higher bar than professionals working around data gaps, since a model inherits gaps and fills them with guesses rather than flagging them.
Trust varies by task. Informational tasks earned the most trust, with 55 percent highly trusting AI to summarize diligence documents and 53 percent to screen opportunities for baseline fit. Judgment tasks closest to valuation earned the least, with 35 percent highly trusting AI to score and rank opportunities and 36 percent to recommend comparable properties. Underwriting had the lowest adoption of any lifecycle stage at 48 percent, yet 23 percent named it as the stage where AI has delivered the most measurable value, second only to portfolio management at 31 percent.
Dealpath frames the outlook as a choice between two costs: the recurring verification tax on outputs built on fragmented data, and a one-time data toll to structure, connect and govern deal data. The report also cites 71 percent of respondents using general-purpose large language models directly, arguing that a firm's competitive edge will therefore lie in proprietary deal flow and recorded investment decisions. Dealpath's 2025 State of AI Readiness survey of 100 institutional investors, cited for comparison, found 61 percent expected AI to deliver faster deals; in 2026 only 35 percent named deal speed among the most-improved outcomes. The report concludes that the return on AI in 2027 will depend more on the context firms provide to models than on which model they use.
Go deeper than the headline
Research every firm in this story, track the ones that matter and get the daily briefing — free for 7 days.