AI in Real Estate Statistics 2026
By Axis Intelligence Research
Co-author: Sarah Mitchell | Last updated: October 27, 2026 | License: CC BY 4.0
Real estate has become one of the AI-forward sectors of the European economy. Axis Intelligence Research’s Property AI Lead Ratio (PALR) reached 124.1 in 2025: 24.76% of EU real estate firms used AI, against 19.95% of all businesses, per Eurostat. In 2021 the same ratio stood at 92.9.
Quick Answer
In the United States, 79% of Realtors use AI at least occasionally and 48% use it weekly or daily, according to the National Association of Realtors’ (NAR) 2026 Technology Report. At the firm level, adoption is far lower: 29.0% of U.S. real estate firms used AI in the latest Census Bureau survey, collected October 5–18, 2026. In Europe, 24.76% of real estate firms used AI in 2025, per Eurostat.
Key Findings
- Axis Intelligence Research’s Property AI Lead Ratio (PALR) for EU real estate rose from 92.9 in 2021 to 124.1 in 2025, meaning real estate firms moved from trailing the economy on AI to adopting it about a quarter faster, based on Eurostat data.
- Only 10 of the 22 EU member states with published data have a PALR above 100, and the median member state scores 98.0, according to Axis Intelligence Research: the EU-wide lead is carried by large markets such as Germany, France and Spain.
- 61.8% of EU real estate firms that use AI apply text mining, while only 18.5% use machine learning for data analysis, against 25.6% across the economy, per an Axis Intelligence Research calculation from Eurostat 2025 data.
- The share of U.S. real estate firms using AI rose from 21.0% to 29.0% between the Census Bureau’s first and latest readings under its current question (November 2025 to October 2026), ranking real estate 6th of 19 sectors, according to Axis Intelligence Research analysis of Business Trends and Outlook Survey data.
- As of October 26, 2026, 18 U.S. jurisdictions have enacted restrictions on algorithmic rent pricing, and 15 of them are in force and enforceable, according to an Axis Intelligence Research count.
How Many Real Estate Agents Use AI in 2026?
Nearly four in five. In NAR’s 2026 Realtors Technology Report, released September 22, 2026, 23% of Realtors said they use AI daily, 25% weekly and 31% occasionally. That adds up to 79% using AI at some frequency, an Axis Intelligence Research sum. Another 9% are curious but have not tried it, and 12% neither use AI nor plan to.
How often Realtors use AI, 2026
| AI use frequency | Share of Realtors | Source |
|---|---|---|
| Daily | 23% | NAR, 2026 Technology Report release |
| Weekly | 25% | NAR |
| Experimenting occasionally | 31% | NAR |
| Curious, haven’t tried | 9% | NAR |
| Not using AI, no plans to | 12% | NAR |
| Any frequency (Axis sum) | 79% | Axis Intelligence Research, from NAR |
NAR’s 2025 survey counted 32% of agents who had not yet tried AI. The 2026 release compares that with this year’s 12% “no plans” group, but the two categories are not worded the same way, so the year-over-year drop is approximate.
How many Realtors is that in absolute numbers?
NAR reported 1,438,569 members in June 2026, down from 1,463,352 a year earlier. Applying the survey shares to that membership, Axis Intelligence Research estimates:
| Group | Share (NAR 2026 survey) | Estimated members | Method |
|---|---|---|---|
| Use AI at any frequency | 79% | 1,136,470 | 1,438,569 × 79% |
| Use AI weekly or daily | 48% | 690,513 | 1,438,569 × 48% |
| Use AI daily | 23% | 330,871 | 1,438,569 × 23% |
| Not using AI, no plans | 12% | 172,628 | 1,438,569 × 12% |
These are estimates. They assume the survey’s shares hold across the whole membership; NAR’s technology survey is a voluntary online questionnaire, and agents who already use technology may be more likely to answer one.
What do real estate agents use AI for?
Marketing copy, overwhelmingly. Among Realtors who use AI, 75% write listing descriptions with it, 56% create social media posts and 52% draft emails and follow-ups, according to NAR. Work closer to the transaction itself is less common: 30% use AI for market summaries and 27% for document review and summarizing.
| AI application (among Realtors who use AI) | Share | Source |
|---|---|---|
| Writing listing descriptions | 75% | NAR 2026 |
| Social media posts | 56% | NAR 2026 |
| Emails and follow-ups | 52% | NAR 2026 |
| Market summaries | 30% | NAR 2026 |
| Marketing content in a personal tone | 30% | NAR 2026 |
| Document review and summarizing | 27% | NAR 2026 |
Sarah Mitchell, analysis: The list reads like a marketing department, not a brokerage. Listing remarks, social posts and nurture emails are the parts of the job where a bad draft costs nothing: the agent reads it, fixes the square footage, and posts. Pricing a listing or reviewing a purchase agreement is where a confident wrong answer costs a commission or a client. The 27% doing document review is the number to watch. If it climbs, AI has moved from the agent’s marketing stack into the transaction file, and that is where brokers will start writing policy.
Which tools do Realtors use alongside AI?
AI-generated content tools such as ChatGPT, Copilot and Gemini are used by 41% of Realtors, which puts AI behind the multiple listing service (MLS, 96%), e-signature (79%), showing scheduling (68%), comparative market analysis (CMA) and pricing tools (59%), drone photography (48%) and customer relationship management (CRM) software (46%), per NAR.
Results are starting to register: 55% of Realtors say AI has had a positive effect on their business, according to NAR, and 40% say clients responded very positively to the use of technology in the buying and selling process, while 37% say clients found it helpful but had some reservations.
The adoption motive has hardened. Saving time is the primary reason for 81% of Realtors, up from 66% in 2025; improving the client experience rose to 71% from 64%. The obstacles are the learning curve (63%) and cost (59%). On spending, 36% of Realtors put $50 to $250 a month into technology and 22% spend more than $500.
AI Adoption in Real Estate: Agents vs Firms vs Workers
Three U.S. surveys measure AI in real estate, and they land in very different places because they measure different things.
| Survey | Who answers | AI measure | Real estate | All industries | Period | Source |
|---|---|---|---|---|---|---|
| Census Business Trends and Outlook Survey (BTOS), latest | Firms | Firm used AI in any business function in the last two weeks | 29.0% | — | Collected Oct 5–18, 2026 | U.S. Census Bureau, BTOS data |
| Census BTOS, as summarized by the Fed | Firms | Same question | 24% | 18% | Real estate: average of four surveys through year-end 2025; all industries: Dec 2025 | Federal Reserve Board, FEDS Notes |
| Real-Time Population Survey (RPS) | Individual workers | Uses generative AI | 58% | 41% | November 2025 | Federal Reserve Board |
| NAR Technology Survey | Realtors | Uses AI at any frequency | 79% | — | 2026 survey | NAR |
Axis Intelligence Research does not merge these three readings into a single “real estate AI adoption” figure. A firm-level yes/no, a worker-level generative AI question and a trade association’s member survey have different denominators. Averaging them would produce a number that describes nobody.
What can be compared is each survey against its own all-industry rate. In the RPS, real estate workers use generative AI at 1.41 times the labor-force rate (58% ÷ 41% = ratio of 141.5), an Axis Intelligence Research calculation from a single survey wave.
For context on the firm side, the U.S. Census Bureau measured 19.8% of all U.S. businesses using AI in the period ending May 3, 2026, with the Information sector at 39.7% and Finance and Insurance at 33.9%. Among firms with 250 or more employees, the rate was 37%.
Sarah Mitchell, analysis: The gap between 79% of agents and 29% of firms is not a contradiction; the surveys count different things. When an agent pays for a ChatGPT subscription personally, that use shows up in NAR’s survey and never in the brokerage’s answer to the Census Bureau. That has a practical consequence for brokers: some of the AI touching client data in residential transactions is procured by the agent, not the firm that holds the license and the errors-and-omissions policy. A brokerage AI policy that only covers firm-purchased tools covers part of the exposure.
How fast are U.S. real estate firms adopting AI?
Steadily, by the Census Bureau’s measure. In the Business Trends and Outlook Survey, the share of real estate and rental and leasing firms (NAICS 53) that used AI in the previous two weeks rose from 21.0% in the first reading under the current question (collected November 17–30, 2025) to 29.0% in the latest (collected October 5–18, 2026, published October 22, 2026). The latest estimate carries a standard error of 1.03 percentage points.
Single readings move around: the series peaked at 31.9% in late August 2026. Averages are steadier. Axis Intelligence Research calculates that the first six readings (November 2025 to early February 2026) average 23.5%, and the latest six (late July to mid-October 2026) average 29.7%, a gain of 6.2 points. Another 35.4% of real estate firms expect to use AI within the next six months.
| Census BTOS reading (collection period) | Real estate firms using AI | Expect to use AI in six months |
|---|---|---|
| Nov 17–30, 2025 (first reading) | 21.0% | 25.1% |
| Dec 15–28, 2025 | 26.4% | 29.5% |
| Apr 6–19, 2026 | 30.0% | 33.5% |
| Aug 24 – Sep 6, 2026 (peak) | 31.9% | 37.7% |
| Oct 5–18, 2026 (latest) | 29.0% | 35.4% |
| Average, first six readings | 23.5% | — |
| Average, latest six readings | 29.7% | — |
Source: U.S. Census Bureau, BTOS sector data (Sector.xlsx, published October 22, 2026); averages by Axis Intelligence Research. All 24 readings are in the dataset.
Where does real estate rank among U.S. industries?
Sixth of 19. In the October 2026 reading, real estate (29.0%) trails Information (46.4%), Professional, Scientific and Technical Services (43.8%), Finance and Insurance (40.5%), Management of Companies (36.4%) and Educational Services (33.4%), and leads Health Care (27.6%), Manufacturing (22.9%) and Retail (18.5%). Axis Intelligence Research calculates that real estate firms use AI at 1.36 times the rate of the median U.S. sector (21.4%).
| Sector (NAICS) | Firms using AI, Oct 2026 |
|---|---|
| Information (51) | 46.4% |
| Professional, Scientific and Technical Services (54) | 43.8% |
| Finance and Insurance (52) | 40.5% |
| Management of Companies and Enterprises (55) | 36.4% |
| Educational Services (61) | 33.4% |
| Real Estate and Rental and Leasing (53) | 29.0% |
| Health Care and Social Assistance (62) | 27.6% |
| Manufacturing (31–33) | 22.9% |
| Administrative and Support Services (56) | 22.7% |
| Wholesale Trade (42) | 21.4% |
| Arts, Entertainment and Recreation (71) | 21.4% |
| Retail Trade (44–45) | 18.5% |
| Mining (21) | 16.8% |
| Utilities (22) | 16.7% |
| Construction (23) | 16.1% |
| Other Services (81) | 14.6% |
| Transportation and Warehousing (48–49) | 13.5% |
| Accommodation and Food Services (72) | 10.7% |
| Agriculture (11) | 6.0% |
Source: U.S. Census Bureau, BTOS sector data, collection period October 5–18, 2026.
Sarah Mitchell, analysis: Sixth place puts real estate in the company of the professional-services sectors, not the field trades it is often grouped with. That fits what the transaction looks like: offer letters, disclosures, lease packets, owner statements. The firms below real estate in this table move goods or people; the ones above it move documents and advice. The next readings will show whether the climb from the low twenties continues or settles near 30%, which is where the series has hovered since August.
For lender-side AI, including origination costs and the Fannie Mae and Freddie Mac AI governance rules, see our AI in mortgage lending statistics. Housing-finance volumes are tracked in mortgage statistics.
Is Real Estate Ahead of Other Industries on AI?
In Europe, yes, and by a widening margin. Axis Intelligence Research measures the gap with the Property AI Lead Ratio (PALR): the share of real estate firms using AI divided by the share of all businesses using AI in the same survey, country and year, multiplied by 100. A PALR of 100 means real estate adopts AI exactly as fast as the average business; above 100, it adopts faster.
Both inputs come from Eurostat’s enterprise survey on ICT use (dataset isoc_eb_ain2, firms with 10 or more employees), so the ratio compares like with like.
EU real estate AI adoption vs the whole economy, 2021–2025
| Year | Real estate firms using AI | All businesses using AI | Gap (pp) | PALR | Source |
|---|---|---|---|---|---|
| 2021 | 7.11% | 7.65% | −0.54 | 92.9 | Eurostat; PALR by Axis Intelligence Research |
| 2023 | 8.50% | 8.06% | +0.44 | 105.5 | Eurostat; Axis Intelligence Research |
| 2024 | 15.45% | 13.48% | +1.97 | 114.6 | Eurostat; Axis Intelligence Research |
| 2025 | 24.76% | 19.95% | +4.81 | 124.1 | Eurostat; Axis Intelligence Research |
EU real estate firms multiplied their AI adoption 3.48 times between 2021 and 2025, a gain of 17.65 percentage points; the economy as a whole multiplied its rate 2.61 times (+12.30 points), according to Axis Intelligence Research calculations from Eurostat data.
One data note: Eurostat’s Statistics Explained article, extracted in December 2025, gives 24.82% for real estate in 2025. The current database value, retrieved on October 26, 2026, is 24.76%. This page uses the database value.
Sarah Mitchell, analysis: Real estate crossed the line in 2023, the year generative AI arrived in every browser. That timing matters. A property manager’s day is leases, tenant emails, maintenance tickets and owner reports, all of it text, so the sector got more out of a writing tool than a factory or a haulier did. Compare the AI in transportation statistics: the same Eurostat survey puts transport firms at 11.15%, less than half the real estate rate. The sector that runs on documents adopted the technology built for documents.
Which Countries’ Real Estate Firms Use AI the Most?
Sweden (56.44%) and Finland (50.93%) lead Europe, with more than half of real estate firms using at least one AI technology in 2025. Spain (34.39%), Germany (29.06%), Croatia (27.47%) and France (26.25%) follow among EU member states, per Eurostat. Romania (0.58%), Poland (4.09%) and Hungary (4.69%) have the lowest rates.
Real estate AI adoption and PALR by country, 2025
| Country | Real estate firms using AI | All businesses using AI | PALR |
|---|---|---|---|
| Croatia | 27.47% | 15.19% | 180.8 |
| Spain | 34.39% | 20.27% | 169.7 |
| Sweden | 56.44% | 35.04% | 161.1 |
| Greece | 12.91% | 8.93% | 144.6 |
| France | 26.25% | 18.16% | 144.5 |
| Finland | 50.93% | 37.82% | 134.7 |
| Portugal | 14.80% | 11.54% | 128.2 |
| Bulgaria | 10.58% | 8.55% | 123.7 |
| Germany | 29.06% | 25.97% | 111.9 |
| Slovenia | 23.15% | 21.61% | 107.1 |
| Estonia | 23.25% | 23.40% | 99.4 |
| Cyprus | 8.95% | 9.27% | 96.5 |
| Ireland | 18.53% | 19.64% | 94.3 |
| Latvia | 11.45% | 12.21% | 93.8 |
| Malta | 19.79% | 21.51% | 92.0 |
| Czechia | 15.86% | 17.60% | 90.1 |
| Lithuania | 19.03% | 21.30% | 89.3 |
| Italy | 13.96% | 16.40% | 85.1 |
| Poland | 4.09% | 8.36% | 48.9 |
| Slovakia | 8.55% | 18.00% | 47.5 |
| Hungary | 4.69% | 10.37% | 45.2 |
| Romania | 0.58% | 5.21% | 11.1 |
| EU-27 | 24.76% | 19.95% | 124.1 |
| Outside the EU: Albania | 21.28% | 8.99% | 236.7 |
| Norway | 36.25% | 28.89% | 125.5 |
| Türkiye | 6.46% | 7.41% | 87.2 |
| Serbia | 5.53% | 10.12% | 54.6 |
Source: Eurostat isoc_eb_ain2, 2025, enterprises with 10+ employees using at least one AI technology; PALR calculated by Axis Intelligence Research. Belgium, Denmark, the Netherlands and Austria have real estate values flagged as low reliability and unpublished; Luxembourg’s value is confidential.
Why is the EU lead driven by a few large markets?
The EU-wide PALR of 124.1 hides a split. Of the 22 member states with a published real estate value, only 10 score above 100, and the median member state sits at 98.0, below parity, per Axis Intelligence Research. The EU aggregate is weighted by the number of firms in each country, so Germany, France and Spain, all with PALR readings above 100 and large real estate sectors, lift the European figure above what the typical member state shows.
Sarah Mitchell, analysis: The country table is a reminder that “real estate” means different businesses in different markets. Where agencies sell to foreign buyers or run holiday lets, a listing has to exist in several languages and guest messages never stop, and text generation pays for itself on the first week of use. Eurostat’s data cannot say which business model drives each country’s score, so treat that as a hypothesis to test, not a finding. What the data does show is the other end of the table: in Poland, Hungary and Slovakia, real estate firms adopt AI at less than half the pace of their own national economies, so the gap there sits inside the sector, not in the availability of the tools.
What Do Real Estate Firms Actually Use AI For?
Mostly language. Among EU real estate firms in 2025, text mining (15.31% of all firms) is the most common AI technology, followed by natural language generation (10.15%) and image, video or audio generation (9.75%), per Eurostat. Machine learning for data analysis, the technology behind automated valuation and rent forecasting, reaches only 4.58%.
AI technologies in EU real estate firms vs all businesses, 2025
| AI technology | Real estate | All businesses | Source |
|---|---|---|---|
| Text mining (analysis of written language) | 15.31% | 11.75% | Eurostat |
| Natural language or code generation | 10.15% | 8.76% | Eurostat |
| Image, video or audio generation | 9.75% | 9.55% | Eurostat |
| Speech recognition | 8.81% | 7.22% | Eurostat |
| AI-based robotic process automation | 5.76% | 5.35% | Eurostat |
| Machine learning for data analysis | 4.58% | 5.10% | Eurostat |
| Image recognition | 3.45% | 3.78% | Eurostat |
| Autonomous robots, vehicles or drones | 1.12% | 1.39% | Eurostat |
Values are shares of all enterprises in each group, not of AI users.
Axis Intelligence Research calculates that 61.8% of AI-using real estate firms use text mining (15.31 ÷ 24.76), against 58.9% across the economy. Only 18.5% use machine learning for data analysis (4.58 ÷ 24.76), compared with 25.6% of AI-using firms overall. Real estate’s AI lead is a language lead.
Which business functions get the AI?
| Purpose of AI use | Real estate | All businesses |
|---|---|---|
| Marketing or sales | 9.00% | 6.92% |
| Business administration or management | 7.81% | 6.19% |
| Accounting, controlling or finance | 6.59% | 4.61% |
| ICT security | 4.67% | 3.94% |
| Production processes | 4.32% | 4.14% |
| R&D or innovation | 2.46% | 3.80% |
| Logistics | 0.66% | 1.21% |
Source: Eurostat isoc_eb_ain2, EU-27, 2025.
Real estate firms are 1.30 times as likely as the average business to use AI for marketing or sales (9.00% vs 6.92%), per Axis Intelligence Research. That matches the U.S. agent data: in both markets, AI shows up first in the listing and the lead, not in the valuation model.
Sarah Mitchell, analysis: The vendor pitch for property AI is a model that prices assets, predicts vacancy and spots the next submarket. The adoption data says the industry bought a writer first. Machine learning for analysis sits below the economy average, which means the predictive layer is concentrated in a smaller group of large owners, portals and data providers, while the long tail of agencies and managers uses AI to answer emails faster. Those are two different markets, and a single “AI in real estate” growth figure blends them.
Why don’t more real estate firms use AI?
Lack of expertise is the barrier most cited by EU real estate firms that don’t use AI: 9.88% of all real estate firms name it, ahead of data-protection and privacy concerns (8.94%) and unclear legal consequences (8.40%), per Eurostat. Cost is cited by only 4.08%, and 1.81% say AI is not useful to them. Another 14.52% of real estate firms have considered using AI, against 11.10% of all businesses.
Privacy and legal concerns rank higher in real estate than across the economy, where they are cited by 5.82% and 5.92% of firms. Real estate firms handle tenant files, identity documents and income proofs, which may help explain the extra caution.
How Is AI Used in Commercial Real Estate?
Through pilots. In JLL’s 2025 Global Real Estate Technology Survey of more than 1,500 senior decision-makers across 16 markets, 88% of investors, owners and landlords and 92% of corporate occupiers said they had started AI pilots, according to JLL’s October 28, 2025 release. Only 5% of occupiers reported achieving all their AI program goals.
| Commercial real estate AI metric | Value | Source |
|---|---|---|
| Investors, owners and landlords piloting AI | 88% | JLL, 2025 |
| Occupiers running corporate real estate AI pilots | 92% | JLL, 2025 |
| Average AI pilots run in parallel | 5 | JLL, 2025 |
| Investors whose tech budgets rose because of AI | 87% | JLL, 2025 |
| Occupiers that achieved all AI goals | 5% | JLL, 2025 |
| Occupiers that achieved two to three goals | 47% | JLL, 2025 |
| AI use cases identified across the CRE value chain | 56 | JLL, 2025 |
JLL sells real estate services and technology, and the survey is its own research. It is also a year old: JLL has not published a 2026 edition, so this is the most recent reading from the series.
Sarah Mitchell, analysis: Ninety-two percent piloting and five percent hitting every goal is the familiar shape of an enterprise technology cycle, and in commercial real estate it has a specific cause. Lease abstraction, operating-expense reconciliation and building data all sit in systems that were never designed to talk to each other. A pilot runs on one clean portfolio; scaling it means cleaning the other forty. The pilot count measures interest. The 5% measures how much of the data plumbing is actually done.
Is It Legal for Landlords to Use AI to Set Rent?
It depends on where the property is and what data the software uses. As of October 26, 2026, Axis Intelligence Research counts 18 U.S. jurisdictions that have enacted restrictions on algorithmic rent-setting: four states and 14 cities or counties. Sixteen are in force, and 15 are currently enforceable, because a federal court has paused New York’s statute.
Most of these laws do not ban pricing software as such. They target tools that pool nonpublic data from competing landlords, such as actual rents, occupancy and lease dates, to recommend rents or occupancy levels. Software working from a landlord’s own data and public listings generally falls outside them.
State laws restricting algorithmic rent pricing
| State | Law | Effective date | Status, Oct 26, 2026 | Source |
|---|---|---|---|---|
| New York | S.7882 (amends the Donnelly Act) | December 2025 | Preliminary injunction granted to RealPage, S.D.N.Y. (Sept. 29, 2026 filings) | Multifamily Dive |
| California | AB 325, Chapter 338 (signed Oct 6, 2025; adds Business and Professions Code §§ 16729 and 16756.1 to the Cartwright Act) | January 1, 2026 | In force | California Legislature (bill text); effective date per Cleary Gottlieb |
| Connecticut | HB 8002 (amends the Connecticut Antitrust Act) | January 1, 2026 | In force | Baker McKenzie |
| New Jersey | FAIR Act (signed July 20, 2026) | July 1, 2027 | Enacted, not yet in force | Greenberg Traurig |
California’s statute defines a “common pricing algorithm” as any methodology, including software, used by two or more persons that uses competitor data to recommend, align, stabilize, set or otherwise influence a price or commercial term. The definition is not limited to rent.
Penalties are not symbolic. Connecticut allows penalties of up to $1,000,000 for corporations and $100,000 for individuals. California’s amended Cartwright Act raised the maximum criminal fine for a corporation to the greater of $6 million or twice the gain or loss.
City and county ordinances
| Jurisdiction | Effective |
|---|---|
| San Francisco, CA | October 2024 |
| Philadelphia, PA | February 2025 |
| Providence, RI | May 2025 |
| San Diego, CA | June 2025 |
| Jersey City, NJ | June 2025 |
| Hoboken, NJ | July 2025 |
| Seattle, WA | July 2025 |
| King County, WA | October 2025 |
| Berkeley, CA | January 2026 |
| Spokane, WA | January 2026 |
| Portland, OR | February 2026 |
| Minneapolis, MN | March 1, 2026 |
| Santa Ana, CA | April 2, 2026 |
| Rockville, MD | January 1, 2027 |
Source: Morgan Lewis, August 2026, which lists the code citation for each ordinance. Morgan Lewis also reports measures pending in Montgomery County, MD, Evanston, IL and Rhode Island; they are not counted here.
What did the RealPage settlement change?
Under the settlement announced by the U.S. Department of Justice on November 24, 2025, RealPage must stop using competitors’ nonpublic, competitively sensitive data to set prices in live operation, may train its models only on nonpublic data at least 12 months old, and may not use models that determine geographic effects narrower than the state level. A court-appointed monitor oversees compliance.
Sarah Mitchell, analysis: Read together, the settlement and the statutes draw the same line: own data and public comps are fine, a shared pool of competitors’ live lease data is not. That moves the risk from the software vendor to the operator’s configuration. A revenue manager who switches on a feed of peer occupancy can turn a lawful tool into an unlawful one in a few clicks. Expect the next wave of disputes to be about data lineage, which inputs went into a recommendation, rather than about whether a landlord used software at all.
For a broader view of AI rules worldwide, see Axis Intelligence Research’s AI regulation by country.
Methodology
Axis Intelligence Research assembled this dataset on October 26, 2026. Every figure was retrieved from the organization that produced it, or, for state and local laws, from law-firm publications that cite the underlying statute or code section.
European data. Country and sector values come from Eurostat’s isoc_eb_ain2 dataset, queried through Eurostat’s dissemination API for real estate activities (NACE section L) and for all activities excluding agriculture, mining and the financial sector (aggregate C10-S951_X_K), for enterprises with 10 or more persons employed. Technology, purpose and barrier values are shares of all enterprises in each group, not of AI users.
Property AI Lead Ratio (PALR). PALR = (share of real estate firms using at least one AI technology ÷ share of all firms using at least one AI technology) × 100, with both shares from the same survey, country and year. PALR compares real estate with its own national economy; it does not rank countries by absolute adoption. Sweden’s real estate firms use AI at a higher rate than Croatia’s (56.44% vs 27.47%), even though Croatia’s PALR is higher. The EU median is the mean of the 11th and 12th values among the 22 member states with a published real estate figure.
Derived shares. Shares of AI users (for example, text mining at 61.8%) divide a technology value by the any-AI value for the same group. Estimated Realtor counts multiply NAR’s June 2026 membership by the 2026 survey shares; they assume the survey is representative of the full membership.
U.S. Census BTOS. The real estate series and the sector ranking come from the Census Bureau’s BTOS sector file (Sector.xlsx, Response Estimates, question 7), downloaded October 26, 2026, covering every biweekly reading since the question was revised in November 2025 (sample periods 202524 to 202621). Values are the share of firms answering “yes”; “no” and “do not know” make up the rest. The file’s question text now reads “Super Intelligence or SI (formerly referred to as Artificial Intelligence or AI)” and lists AI-powered data analysis, generative tools for marketing, design or coding, virtual agents for customer service, forecasting or pricing, and computer vision as examples; this page reports it as AI use. The sector file’s residual category (code XX) is excluded from the ranking and the median; it is not the national rate. The median is taken across the 19 named sectors.
U.S. data, other. Worker readings and the year-end 2025 firm summary come from the Federal Reserve Board’s April 3, 2026 FEDS Note, which reports Census Business Trends and Outlook Survey and Real-Time Population Survey results for real estate. The Census Bureau’s May 26, 2026 story supplies the national rate for the period ending May 3, 2026. Agent-level figures come from NAR’s 2026 Technology Report release and report page.
Regulation count. The 18 jurisdictions are four state statutes (New York, California, Connecticut, New Jersey) and the 14 local ordinances listed by Morgan Lewis in August 2026. “In force” means the effective date had passed by October 26, 2026; “enforceable” further excludes New York, where a preliminary injunction applies.
What is not here. Market-size estimates for “AI in real estate” are excluded. The figures in circulation range from a few billion dollars to several hundred billion for the same year, and none of the reports we reviewed discloses a method that would allow its number to be reproduced.
About This Dataset
The file ai-in-real-estate-statistics.csv holds 328 rows: EU real estate and economy-wide AI adoption by year and country, PALR readings, AI technologies, purposes and barriers, U.S. firm adoption (24 biweekly Census readings and a 19-sector ranking), worker and Realtor adoption, NAR use cases and technology use, JLL commercial real estate figures, and 18 algorithmic rent-pricing laws with effective dates. Each row carries value, unit, as-of date, reference period, geography, segment, source organization, source document, source URL, retrieval date, a primary-source flag, an Axis-calculated flag, the formula for every calculated row and a data-quality flag where one applies.
Creator and publisher: Axis Intelligence Research. License: CC BY 4.0, free to reuse with attribution. Citation line: Axis Intelligence Research, AI in Real Estate Statistics 2026, 2026.
How to Cite This Page
APA
Axis Intelligence Research, & Mitchell, S. (2026, October 27). AI in real estate statistics 2026: Agent adoption, firm use by country and the rent-algorithm laws [Dataset]. Axis Intelligence. https://axis-intelligence.com/ai-in-real-estate-statistics/
MLA
Axis Intelligence Research, and Sarah Mitchell. "AI in Real Estate Statistics 2026: Agent Adoption, Firm Use by Country and the Rent-Algorithm Laws." Axis Intelligence, 27 Oct. 2026, axis-intelligence.com/ai-in-real-estate-statistics/.
Chicago
Axis Intelligence Research, and Sarah Mitchell. "AI in Real Estate Statistics 2026: Agent Adoption, Firm Use by Country and the Rent-Algorithm Laws." Axis Intelligence, October 27, 2026. https://axis-intelligence.com/ai-in-real-estate-statistics/.
AI in Real Estate: Questions Agents, Brokers and Landlords Ask
Can landlords still use rent-pricing software in 2026?
Yes, with limits. As of October 26, 2026, 15 U.S. jurisdictions enforce restrictions on algorithmic rent-setting, mostly aimed at software that pools nonpublic data from competing landlords, per an Axis Intelligence Research count. Tools that price from a landlord’s own data and public listings generally fall outside these laws.
Is New York’s ban on algorithmic rent pricing in effect?
Not currently. New York’s law took effect in December 2025, but RealPage won a preliminary injunction in the Southern District of New York, according to September 29, 2026 court filings reported by Multifamily Dive. The ban is paused while the court weighs RealPage’s First Amendment claims.
What share of Realtors use ChatGPT or other AI writing tools?
41% of Realtors use AI-generated content tools such as ChatGPT, Copilot or Gemini, according to NAR’s 2026 Technology Report. Counting any AI use, 79% of Realtors use AI at least occasionally, and 75% of those who use it write listing descriptions with it.
Do real estate brokerages use AI as much as their agents?
No. The Census Bureau found 29.0% of U.S. real estate firms used AI in its October 5–18, 2026 survey, while 79% of Realtors use AI at some frequency in NAR’s 2026 survey. Much agent AI use runs on personal subscriptions that a firm-level survey does not capture.
Which European country’s real estate firms use AI the most?
Sweden, where 56.44% of real estate firms used AI in 2025, followed by Finland at 50.93%, per Eurostat. Relative to its own economy, Croatia leads the EU with a Property AI Lead Ratio of 180.8, according to Axis Intelligence Research.
Is AI used to value homes and price listings?
Less than marketing suggests. Among EU real estate firms, 4.58% use machine learning for data analysis, against 15.31% using text mining, per Eurostat 2025. In the U.S., 30% of AI-using Realtors use it for market summaries, compared with 75% for listing descriptions, according to NAR.
What happened in the DOJ case against RealPage?
The Justice Department announced a settlement on November 24, 2025. RealPage must stop using competitors’ nonpublic data in live pricing, train only on nonpublic data at least 12 months old, avoid models narrower than state level, and accept a court-appointed monitor, per the DOJ.
Are commercial real estate firms getting results from AI?
Few have reached their goals. JLL’s 2025 survey found 92% of corporate occupiers running AI pilots but only 5% achieving all their AI program goals; 47% had achieved two to three goals. Investors run an average of five pilots at once.
Related Axis Intelligence Research pages: AI adoption by industry · AI in mortgage lending statistics · AI in insurance statistics · AI in transportation statistics · Corporate AI adoption statistics · AI job losses by industry · Mortgage statistics