1 in 3
Homebuyers who say a significant AI listing error would reduce their trust in the platform.
68%
The percentage of homebuyers who want clear notification when AI is involved in outputs.
44%
Buyers who would pay more for a human to verify AI outputs.
Tens of millions of buyers welcomed it in, fast and without much resistance. What they want from it next is the harder, more interesting question.
Fifty-five percent of homebuyers now use generative AI tools at least once a month. Three-quarters already assume AI is running somewhere inside the homebuying process — in property search, in valuations, in the rate quotes that appear on their screens. By those measures, the industry's AI moment has arrived in a big way.
But peel back the curtain to look beyond the surface-level acceptance, the numbers tell a different story.
Introduction
.svg)
Among prospective buyers in the United States, trust in AI tools to help find a home dropped nearly in half, from 30% in 2025 to 16% today. Preference for working with human professionals has risen across every major task: finding a mortgage, securing homeowners insurance, and navigating legal paperwork. Sixty-eight percent of buyers say they would manually verify every detail, or a significant amount, of anything an AI tool provides them in a housing context.
Buyers who once approached AI as a novelty now assess it the way they assess any other infrastructure: by whether its answers are accurate, transparent, and recoverable when it goes wrong. Familiarity, it turns out, creates standards rather than complacency.
The pattern holds across the four geographies in which we conducted our survey, though with meaningful variation. U.S. and Canadian buyers are more open to AI involvement in automated valuations and rate decisions than their United Kingdom and Australian counterparts. Millennials lead AI tool adoption and also lead willingness to pay for human verification of AI outputs. Gen Z is the most likely generation to gain confidence from AI's involvement in a transaction, and also among the most likely to invest in verifying what it produces. This finding comes on the heels of a Cotality survey last year that showed Gen Z losing confidence as they progressed into the buying journey. It appears AI can help ease that anxiety.
Although Baby Boomers report the most distrust of AI details, younger generations — especially Millennials — are more likely to invest in AI verification from expert professionals.
Generational reliance on human verification of AI outputs
Data source: Cotality survey, 2025
Technology has moved faster than trust in it has followed. AI adoption happened in the open; the trust deficit accumulated in missed moments. This Cotality report measures those gaps and opportunities precisely, by sector, by geography, and by generation.
The starting point is clear: AI is now being held to a higher standard, and closing the trust gap is now the defining market imperative entering 2026.
A 21-year-old recent buyer in the U.S. captured the sentiment plainly: she didn't want to take a risk when she had “in-person real individuals that have years of experience” to lean on. She was not so much dismissing AI as pricing the cost of being wrong.
What this report covers
Cotality's Q1 2026 housing trends research surveyed homebuyers in the U.S., Canada, the U.K., and Australia between January 29 and February 9, 2026. The sample split between recent buyers (purchased within five years) and prospective buyers (planning to purchase within two to five years), with generational coverage across Gen Z, Millennials, Gen X, and Baby Boomers.
The report is organized around six findings that have direct operational implications for lenders, brokers, insurers, regulators, and the wider property industry.
How is AI changing the housing market?
Overarching findings from a survey of recent and prospective homebuyers across the U.S., Canada, U.K., and Australia in Q1 2026.
- Lenders and brokers: Trust in AI tools is softening in a key segment — many ask about bias in automated valuation models. Your response affects customer retention measurably, and in a consistent direction.
- Insurers: You are already perceived as the heaviest users of invisible, background AI. That perception carries the largest share of the trust burden.
- Real estate professionals: Buyers are arriving better informed but less confident. The value of professional guidance when it comes to AI in real estate has shifted in terms of what it needs to deliver.
- Legal and compliance: Consumer expectation is pushing toward norms on disclosure and contestability. The data supports that direction with unusual clarity.
Disclosure starts here
How AI is reshaping the housing market
Industry implications
- Lenders and brokers: Standardize disclosure language across sites, broker portals, and customer communications. Inconsistency between channels reads as concealment.
- Insurers: Place disclosure at the precise moments when AI affects eligibility, coverage, or price. General statements about using data are not sufficient — buyers want to know when and how AI shaped the specific number in front of them.
- Legal and compliance: Draft disclosure language that works simultaneously as a consumer communication and as a defensible audit document — the two functions reinforce each other.
- Regulators: The data supports baseline rules and common definitions across the industry. One-off labeling approaches will produce a patchwork that satisfies no one, and older buyers are pushing for legal requirements — that pressure is measurable and growing.
- Real estate professionals: Align claims and disclosures in sales materials with buyer expectations that AI influence is flagged — particularly where AI-generated valuations or neighborhood assessments appear in listing materials.
Like most disruptive innovations, AI has shifted quickly from novelty to expectation. Three in four buyers say they think it is playing a role in the homebuying process, both through visible interactive tools and the invisible backend algorithms working underneath. And it is easy to see why. Property search sites now serve up precise recommendations in seconds, and 86% of buyers assume that is AI at work. Prequalification mortgage rates appear on the fly. Insurance quotes arrive at a speed once impossible through human underwriting alone.
It is little surprise, then, that more than 80% of buyers attribute this pace to artificial intelligence. Even government housing bodies, hardly known for cutting-edge automation, are assumed to use AI by 74% of buyers. The assumption that AI is already powering much of today’s real estate industry is broad, and shared across generations.
But the mental model buyers carry on AI involvement is more nuanced than a simple yes or no. They distinguish between two roles: foreground AI — tools they interact with directly for search, comparison, and valuation — and background AI that makes decisions about them, including insurance underwriting and lender risk scoring. Property websites score highest for visible, foreground use. Insurers take the top position for invisible, data-driven use, where algorithms act for consumers without their direct engagement.
That distinction carries real commercial weight. Being perceived as an AI-driven process but without clear insight into how it works places insurers at the sharpest end of disclosure scrutiny — and likely first in line should regulatory norms around transparency tighten.
A 32-year-old Millennial future buyer in the UK clearly described the overarching preference of survey respondents for visible AI involvement: she wanted to actively engage in research and decision-making, with the ability to explore different options, ask questions, and receive personalized insights — specifically because having that “sense of control” mattered when the stakes were this high.
"I prefer foreground AI for the homebuying process because it allows me to actively engage in research and decision-making," the 32-year-old Millennial future buyer in the UK said. "I'd feel more confident knowing I'm fully informed and have the ability to explore different options, ask questions, and get personalized insights. It's important for me to have that sense of control, especially for something as big as buying a home."
It is no longer enough to simply say AI is powering insights. Buyers assume AI is present, and 68% say clear notification whenever AI generates a listing, price, or mortgage recommendation is either very important or even a legal requirement. That figure rises sharply with age: 61% of Baby Boomers say it should be legally mandated. Among Gen Z — the cohort most comfortable with AI in every other context — that figure is 25%. But even younger generations are emphatic that notification matters; they simply express it through expectation rather than legislation.
Buyers know AI is in the system. They want to be told precisely when and how it is shaping the specific decision in front of them. Assumed presence and demanded transparency sit comfortably in the same mind — they are not contradictory impulses, they are sequential ones.
Some in the industry have been treating disclosure as a branding question — whether to mention AI as a feature. This misses the point. Buyers have moved on to a governance question: whether they can see, understand, and contest what AI is doing to their options. A 33-year-old Millennial future buyer in the UK put it succinctly: “I trust AI to an extent but would like to make the decisions myself. I think AI is still at the stage where its output needs to be checked manually.”
The geographic picture adds texture. U.S. and Canadian buyers show higher overall acceptance of AI involvement than their UK and Australian counterparts. UK buyers score higher on demanding human involvement across legal assistance and insurance tasks — a market where professional trust hierarchies remain strong. For companies operating across multiple markets, disclosure strategy needs to flex by geography, as well as by product.
The trust cliff
Housing market AI insights
Industry implications
- Lenders and brokers: Raise quality assurance (QA) standards for all AI-touched borrower-facing outputs, especially anything presented as factual — valuations, rate comparisons, eligibility assessments. The error penalty is asymmetric and the trend is accelerating.
- Insurers: People have a lower threshold for errors in AI-powered tools like hazard assessments, mitigation descriptions, or coverage details than their human-powered equivalents. Accuracy in communicating insurance property risk scoring is a trust asset with a measurable value.
- Regulators: Faster trust erosion amplifies the value of remediation standards and accessible complaint pathways. The public is not waiting for regulatory frameworks around AI loan documents accuracy, AI mortgage fraud detection tools, or commercial real estate stress testing models — They are already acting on their expectations.
- Legal and compliance: Strengthen correction processes and record-keeping around automated real estate systems. A 'learning phase' defense will not satisfy buyers who applied zero-tolerance standards long before AI applications in real estate were introduced into the process.
- Real estate professionals: Sales claims about generative AI in real estate need tight sourcing and channel consistency. Discrepancies between online AI real estate market prediction models and in-person information will be noticed, and the forgiveness extended to humans does not transfer to the platforms behind them.
Consider two versions of the same scenario: a significant factual error in a home listing. In one version, it appears on an AI-assisted property website. In the other, the same error is made by a human real estate agent. The outcomes, however, are very different since buyers apply different standards to each scenario.
Seventy percent say a significant AI listing error would reduce their trust in the platform. Sixty percent say the same about a human agent. That ten-point gap — the trust cliff — widens with every generation removed from early digital fluency or familiarity. Baby Boomers show a 14-point gap. Even among Gen Z, where tolerance for technology runs highest in every other context, the gap persists at eight points. The asymmetry is not a function of age; it runs through the entire sample.
Human vs. AI trust gap
Data source: Cotality survey, 2025
The logic buyers apply is consistent. Humans make mistakes because they are human — tired, overloaded, working with incomplete information, and most importantly, capable of relationships. Despite attempts at humanizing AI services, bots are still seen as efficiency tools that deserve no such sympathy. When these tools fall short, the disappointment carries an extra charge of betrayal. As one recent buyer in the UK, a Millennial aged 33, put it: “I expect AI to be pretty much perfect but understand that humans can make mistakes.”
A recent buyer in the U.S. who identified as Gen Z, aged 28, framed the human side of that equation with equal clarity:
What makes the trust cliff operationally significant are the buyers with a zero-trust policy for mistakes. One in three buyers says errors in listing data are unacceptable regardless of source — human or AI. This group is not applying a higher standard to machines; it is applying maximum standards to everything. They are the buyers to worry about after a single significant mistake, and they concentrate in the UK and Australia, and among Baby Boomers and Gen X — the demographics with the highest transaction values and the longest memories.
Homebuyer error tolerance
Data source: Cotality survey, 2025
Among Baby Boomers specifically, the position is blunt. “You can't have error when you are talking about money being spent on a house,” said one 77-year-old future buyer in the U.S. “I have a low tolerance for stupid mistakes.” A 71-year-old recent buyer offered a shorter version: “I would not accept any errors in an important decision.”
The trust cliff is readily apparent in the year-on-year data. U.S. prospective buyers trusting AI tools to help find a home fell 14 points from 2025 to 2026. That decline is not evenly distributed across generations, but the downward trend is apparent across all age groups and rose year-over-year.
Positive sentiment towards AI is declining
Data source: Cotality survey, 2025
AI error in a borrower-facing context is not a training issue or a maturity curve question. Buyers are not interested in calibrating their expectations to where the technology currently sits; they have raised the bar and are asking for technology to follow. A recent Gen X buyer in the U.S., aged 49, described the problem this way: "Trust. AI hallucinates and is not yet reliable.
Why are today’s buyers feeling more discomfort in the homebuying process?
Cotality’s Allie Barefoot talks with Head of Research & Development, Anand Srinivasan.
Accuracy is a trust metric with measurable value. As the gaps in answer quality begin to show and positive sentiment toward AI erodes, data quality will become the foundational framework for companies working to build confidence in their platforms. That’s why Cotality models — which draw from billions of verified records across thousands of sources — are refined through years of applied science to withstand scrutiny and deliver clarity to those who rely on it. Only quality intelligence will translate into technology adoption and customer retention.
