IREM® and AppFolio AI Insights Report: Performance Gains Go to the Most Proficient AI Users
Key Takeaways
AI adoption in real estate management hit 58% in 2026, up from 45% in 2025 and 21% in 2023.
78% of AI users now rely on it daily or almost daily. Adoption has moved from experimentation to habit.
Only 8% of companies have a written AI policy, and 74% of employees have received no formal training.
The most proficient AI users see the strongest financial results: 67% report a positive net operating income (NOI) impact, compared with just 15% of the least mature users.
62% of AI users report a positive impact on resident satisfaction, not just internal productivity.
63% expect AI to complement or reshape their teams. Only 13% expect it to shrink headcount.
Property Managers Just Crossed the 50% Mark on AI Adoption. Governance Hasn't Caught Up.
For the first time since tracking began in 2023, more than half of real estate management professionals say they're using AI in their role. Adoption climbed from 21% in 2023 to 58% today, a nearly threefold increase in three years.
That's the headline, but it's not the whole story.
The deeper finding in AI Insights 2026, produced by IREM® in partnership with AppFolio and based on a survey of 3,662 real estate management professionals, is that adoption and readiness are moving at two different speeds. Companies aren't just behind on strategy. Most have no written AI policy, and most professionals have never received formal training. So while people are moving forward with technology, organizations mostly aren't moving with them.
Everyone's Using AI, but in Different Ways
Last year's report asked a simple question: Are you using AI or not? This year, the AI Maturity Index goes further, scoring respondents across three dimensions: how well they understand AI, how open they are to it, and how skillfully they use it day-to-day.
The results are split into four tiers: Dormant (15%), Curious (40%), Engaged (29%), and Proficient (16%).
The Curious segment, made up of professionals who are experimenting but haven't built AI into daily work, is the largest group in the industry. It's also where the report finds the biggest opportunity. Understanding, openness, and use all rise together, and advancing up the index means growing across all three dimensions at once: understanding AI more deeply, staying open to it, and using it more actively and skillfully over time.
Using AI vs. Getting Results: The Difference Is Depth of Integration
Eighty-five percent of AI users report a productivity gain. That's broad and it's real. But financial outcomes tell a more layered story: Only 48% report a positive impact on NOI or profitability. When AI is used to do more tasks faster instead of changing what the work is for, it creates an efficiency trap, which shows up in the gap between these two numbers. Productivity is the floor, not the ceiling. It's the leading edge of AI's impact, not the end state. Getting people to use AI turns out to be the easy part; getting AI to move the numbers that matter is the harder, unfinished work.
The strongest performers are those using AI most proficiently, with higher gains across NOI, stakeholder outcomes, and other key metrics. As operators move from Dormant to Proficient on the AI Maturity Index, positive NOI impact rises from 15% to 67%, resident satisfaction from 15% to 75%, and owner satisfaction from 13% to 72%. Depth of use, not adoption alone, is what drives performance. What separates Proficient users from Dormant ones isn't just how often they use AI but also how they use it. It's embedded in how operations run, not just how individual tasks get done, with an eye on the outcomes that matter: NOI, resident satisfaction, and owner returns. Maturity here isn't just a usage metric; it's an outcome lens.
As James Scott, IREM’s Innovator-in-Residence and a member of its 2026 Technology Advisory Council, comments in the report: "The shift happens when you stop treating AI as something to consult occasionally and start assigning it a consistent role in your workflow… The difference isn't incremental; it reflects a clear threshold where consistent use starts to compound."
It's the central argument of real estate performance management playing out in the data: The operating model, not the mere presence of AI, determines outcomes. While early-stage users tend to report time saved, the most proficient users report something more consequential: higher resident satisfaction, higher owner/investor satisfaction, and improved business performance (e.g. NOI/profitability, revenue, unit growth).
That resident-side impact matters on its own terms. Sixty-two percent of AI users report a positive impact on resident satisfaction. When property managers have more time to focus on higher-value work, residents get faster answers and more consistent communication, exactly the kind of outcome that builds trust and a sense of being well cared for. Notably, this resident-facing gain is already outpacing harder financial outcomes such as NOI (48% positive impact), which suggests experience improvements for residents are arriving ahead of bottom-line results, not as a byproduct of them.
The real unlock, then, is agentic AI in service of outcomes, not output, and that only works when governance is built in so agents can work self-sufficiently on routine tasks under human monitoring rather than needing manual review at every step. Which brings us to the industry's weakest spot.
The Governance Vacuum Is the Story Underneath the Story
The numbers on impact are strong. The numbers on oversight aren't. Fifty-four percent of companies neither encourage nor discourage AI use, unchanged from last year's survey, even as adoption jumped 13 points over the same period. Only 8% of companies have a written AI policy, and 74% of employees have received no formal training. The resulting policy vacuum pushes the decision, and the risk, onto individual employees with no guidance to fall back on. And direction works when it exists: Employees at companies that actively encourage AI use are twice as likely to use it.
The vacuum shows up sharply around agentic AI, tools that act without a human reviewing every step. Respondents are relatively comfortable letting AI screen applications or draft routine messages. They're far less comfortable letting it near owner and investor communication. That hesitation is rational when there's no governed execution layer — no guardrails, audit trail, or clear accountability to fall back on. Governance is what expands what agentic AI can responsibly take on while keeping humans accountable for the judgment calls that matter most.
Fraud and security concerns reinforce the same instinct toward caution. 98% of respondents have some level of concern about AI-related cybersecurity risk, and 93% are concerned about AI-generated fraudulent applicant documents. The industry sees the risk clearly, which is the first step to managing it.
What This Means for the Year Ahead
The report's own recommendations are a fair summary of where the industry needs to move next: build AI literacy through hands-on training rather than one-off rollouts, put real governance in writing, embed AI into daily workflows instead of one-off tasks, measure service, satisfaction, and NOI rather than time saved alone, and tighten fraud and verification controls as adoption scales.
The throughline across all of it is augmentation, not headcount reduction. Sixty-three percent of respondents expect AI to complement or reshape their teams; only 13% expect it to shrink them. The industry has answered the adoption question. The next one is harder, and worth more. It's not whether you use AI, but how deeply, and whether the gains show up where they count: NOI, resident experience, and owner returns, not just the to-do list.
Read the full AI Insights 2026 Report for the complete data set, including the AI Maturity Index methodology, workflow-by-workflow breakdowns, and expert commentary from IREM's Technology Advisory Council.