AI in Retail Statistics 2026
By Axis Intelligence Research
Co-author: Sarah Mitchell | Last updated: August 1, 2026 | License: CC BY 4.0
Around 14% of U.S. retail trade businesses were using artificial intelligence in any business function as of May 3, 2026, against a national all-sector rate of 19.8%, according to the U.S. Census Bureau’s Business Trends and Outlook Survey. Retail is not leading the AI economy. It is running roughly seven-tenths of the pace of the average American business.
Quick Answer
Retail’s AI adoption rate is 14% of businesses in the United States (Census Bureau, May 3, 2026), well below the 19.8% national rate and 2.84 times below the Information sector’s 39.7%. Axis Intelligence Research finds a Retail AI Adoption Gap of 5.8 percentage points, and — the number that matters more — retail’s own six-month forward intent of 17% still lands 2.8 points below where the rest of the economy already was in May 2026.
Key Findings
- Axis Intelligence Research finds a Retail AI Adoption Gap (RAAG™) of 5.8 percentage points as of May 3, 2026, calculated as the U.S. all-sector AI use rate of 19.8% minus the Retail Trade rate of 14.0%, both from the Census Bureau’s Business Trends and Outlook Survey.
- Axis Intelligence Research finds that retail’s forward AI intent does not close the gap: the roughly 17% of retail businesses expecting to use AI within six months remains 2.8 points below the 19.8% all-sector rate already recorded in May 2026.
- Axis Intelligence Research calculates a Retail AI Capital Divergence Ratio (RACD™) of 6.40× between Amazon and Walmart — Amazon spent 24.35% of quarterly net sales on property and equipment in Q1 2026 versus Walmart’s 3.80% in the quarter ended April 30, 2026, both figures from SEC-filed earnings releases.
- According to the U.S. Census Bureau’s April 2026 working paper CES-26-25, 57% of AI-adopting U.S. firms deploy it in three or fewer business functions, with Sales and Marketing the most common at 52%.
- Eurostat reported that 19.95% of EU enterprises with 10 or more employees used AI technologies in 2025, and that among retail trade enterprises using AI, 48.18% pointed it at marketing or sales.
How Many Retailers Actually Use AI in 2026?
The most defensible answer comes from a survey almost nobody in retail trade press cites: the Census Bureau’s Business Trends and Outlook Survey, which samples about 1.2 million businesses on a biweekly cycle and asks two plain questions — did you use AI in any business function in the past two weeks, and do you expect to in the next six months.
The Census Bureau’s own analysis of the December 14, 2025 to May 3, 2026 collection window found that national AI use hovered between 17% and 20%, settling at 19.8% in the final wave. Retail Trade came in at around 14% current use and about 17% expected use — both below the national line, and both statistically flat across the six-month window.
That flatness is the story. Retail did not stall at 14% because retailers hate technology. It stalled because the AI use case that dominates the vendor pitch decks — copilots for knowledge work — maps poorly onto a sector where the median firm is a single-location store with fewer than twenty employees and no analyst bench to point a copilot at.
AI Use Rate by Sector, United States (May 3, 2026)
| Sector | Current AI use | Expected use, next 6 months | Source |
|---|---|---|---|
| Information | 39.7% | ~42% | U.S. Census Bureau, BTOS |
| Finance and Insurance | 33.9% | ~39% | U.S. Census Bureau, BTOS |
| All sectors (national) | 19.8% | 20%–23% range | U.S. Census Bureau, BTOS |
| Retail Trade (NAICS 44–45) | ~14% | ~17% | U.S. Census Bureau, BTOS |
Data as of the BTOS collection period ending May 3, 2026. Retail and expected-use figures are reported by the Census Bureau as approximate.
AI Use Rate by Firm Size, United States
| Firm size | AI use rate | As of | Source |
|---|---|---|---|
| 250+ employees | 37% | May 3, 2026 | U.S. Census Bureau, BTOS |
| 100–249 employees | 32% | May 3, 2026 | U.S. Census Bureau, BTOS |
| 1–4 employees | Under 20% | May 3, 2026 | U.S. Census Bureau, BTOS |
| All firms, employment-weighted | 32% | Nov 2025–Jan 2026 | U.S. Census Bureau, CES-26-25 |
| All firms, firm-weighted | 18% | Nov 2025–Jan 2026 | U.S. Census Bureau, CES-26-25 |
The firm-weighted versus employment-weighted split is the single most misused pair of numbers in AI adoption coverage. The Census Bureau’s working paper CES-26-25 put firm-level adoption at 18% and employment-weighted adoption at 32% for the same reference period. Both are true. They answer different questions. One counts storefronts; the other counts payroll. A headline that says “a third of businesses use AI” is quoting the payroll number and calling it the storefront number.
Sarah Mitchell, AI & Machine Learning Editor: Every quarter I read a retail keynote claiming AI has become table stakes, and every quarter the federal survey says roughly seven in eight retail businesses did not touch it in the past fortnight. Both statements describe reality — they just describe different retailers. The chains with a data team and a merchandising analytics function are deep in it. The 2,000-store universe is not the retail universe. Census counts the other two million.
The Axis Retail AI Adoption Gap (RAAG™)
Sector adoption rates get quoted in isolation, which hides the thing operators actually need to know: how far behind is retail, is the distance closing, and how fast. Axis Intelligence Research built the Retail AI Adoption Gap to answer that in one reproducible number.
RAAG™ v1.0 — formula and inputs (all inputs U.S. Census Bureau BTOS, snapshot May 3, 2026):
| Component | Formula | Inputs | Reading |
|---|---|---|---|
| Gap-Now | National rate − Retail rate | 19.8 − 14.0 | 5.8 pp |
| Relative Deficit | Gap-Now ÷ National rate | 5.8 ÷ 19.8 | 29.3% |
| Coverage Ratio | Retail rate ÷ National rate | 14.0 ÷ 19.8 | 70.7% |
| Intent Shortfall | National rate now − Retail expected rate | 19.8 − 17.0 | 2.8 pp |
| Leader Spread | Information rate ÷ Retail rate | 39.7 ÷ 14.0 | 2.84× |
Source: U.S. Census Bureau, Business Trends and Outlook Survey, collection period ending May 3, 2026. Calculation by Axis Intelligence Research. Baseline reading; methodology version 1.0. Arithmetic independently verified.
Read the Intent Shortfall first. Retail businesses were asked where they expect to be in six months, and their own answer — about 17% — sits below where the average American business already stood on May 3, 2026. The sector’s optimistic case for late 2026 is the rest of the economy’s spring 2026. Retail plans to add roughly 3 percentage points over six months, a 21.4% relative gain off a low base, and still arrive behind.
Stated limitations. RAAG measures the share of businesses reporting any AI use, not intensity, spend, or outcome. It is firm-weighted, so a Walmart and a single hardware store count once each — which is the point, but it means RAAG will always read lower than any employment- or revenue-weighted measure. BTOS retail and expected-use figures are published as approximations, so RAAG carries their rounding. The index does not capture AI embedded in purchased software that a respondent may not recognise as AI, which almost certainly biases the retail reading downward. This is a baseline reading; there is no prior series to compare it against.
Where Do Retailers Actually Deploy AI?
Two statistical agencies, two questionnaires, two continents, and they land in nearly the same place: AI in commerce goes to the marketing and sales function first, and often only there.
| Deployment site | Rate | Population | Source |
|---|---|---|---|
| Sales and marketing | 52% | U.S. AI-adopting firms, all sectors | U.S. Census Bureau, CES-26-25 |
| Strategy and business development | 45% | U.S. AI-adopting firms, all sectors | U.S. Census Bureau, CES-26-25 |
| Information technology | 41% | U.S. AI-adopting firms, all sectors | U.S. Census Bureau, CES-26-25 |
| Marketing or sales | 48.18% | EU retail trade enterprises using AI | Eurostat, isoc_eb_ain2, 2025 |
| Marketing or sales | 34.70% | EU enterprises using AI, all sectors | Eurostat, isoc_eb_ai, 2025 |
| Logistics | 6.08% | EU enterprises using AI, all sectors | Eurostat, isoc_eb_ai, 2025 |
Axis Intelligence Research notes the convergence: the U.S. figure for sales and marketing among AI adopters (52%) and the EU figure for marketing or sales among AI-using retail trade enterprises (48.18%) sit 3.82 points apart, from independent instruments with different sampling frames and different question wording.
We are deliberately not merging these two figures into a single “global retail AI marketing adoption rate.” The U.S. number covers all sectors; the EU number covers retail trade specifically. The populations differ, the reference periods differ, and averaging them would produce a number that means nothing while looking authoritative. Published side by side, they support a stronger claim than any blend would: the marketing-first pattern survives a change of continent, questionnaire, and statistical agency.
The logistics number deserves more attention than it gets. Across all EU enterprises using AI, only 6.08% pointed it at logistics — the lowest of any purpose Eurostat measured. Retail spends its year fighting shrink, allocation errors, and out-of-stocks, and the function where AI could compound hardest is the function it reaches last. Personalisation engines are easier to buy than a demand-forecasting rebuild, and easier still to put in a board deck.
Sarah Mitchell: The depth data is where the hype breaks. Census found 57% of adopting firms use AI in three or fewer business functions, and 65% confine worker use to three or fewer tasks. That is not transformation. That is a pilot with good PR. The gap between “we deploy AI” and “AI changed how we operate” is the whole 2026 story, and it is measurable now rather than assertable later.
How Much Are Retailers Spending on AI?
Nobody files a line item called “AI.” What large retailers do file is capital expenditure, and in 2026 the split between two companies in the same industry has become extreme enough to function as a measurement in itself.
Amazon’s Q1 2026 earnings release states plainly that a $59.3 billion year-over-year increase in purchases of property and equipment “primarily reflects investments in artificial intelligence.” Trailing-twelve-month free cash flow fell to $1.2 billion from $25.9 billion — a 95% collapse — on the same cause. Walmart, over a comparable quarter, guided fiscal 2027 capital expenditure to approximately 3.5% of net sales and reiterated it.
RACD™ v1.0 — Retail AI Capital Divergence Ratio. Formula: capital expenditure intensity of Amazon ÷ capital expenditure intensity of Walmart, where intensity = purchases of property and equipment ÷ net sales, each company’s most recently reported quarter, both from SEC-filed documents.
| Input | Amazon | Walmart |
|---|---|---|
| Quarter ended | March 31, 2026 | April 30, 2026 |
| Net sales | $181,519M | $175,684M |
| Purchases of property and equipment | $44,203M | $6,684M |
| Capex intensity | 24.35% | 3.80% |
| Source | Amazon Q1 2026 earnings release | Walmart Form 8-K, EX-99.1 |
RACD™ = 24.35 ÷ 3.80 = 6.40×. Spread: 20.55 percentage points. Calculation by Axis Intelligence Research; arithmetic verified.
Stated limitations, and they are load-bearing. Amazon’s capital expenditure is dominated by AWS data-centre construction, not store or fulfilment AI, so RACD measures enterprise-wide reinvestment intensity at two companies that both sell groceries — it is not a retail-only AI spend comparison, and anyone citing it as one is misreading it. The quarters end one month apart. Walmart’s figure is a single quarter against guidance of approximately 3.5% for the full year; on a trailing-twelve-month basis Amazon’s intensity is 19.83% ($147,299M ÷ $742,776M), which against Walmart’s guided 3.5% yields a 5.67× reading. Both readings are disclosed so that a reader can pick the basis they trust.
What the ratio does capture is a structural fact retail boards are living with: one competitor has decided AI infrastructure is the business, and the other has decided it is a tool bought from someone who made that decision. Walmart’s own Q1 FY27 numbers show what the second path buys — global eCommerce up 26%, global advertising up 37%, Walmart Connect excluding VIZIO up 44%, and eCommerce contributing about 530 basis points to U.S. comp sales against roughly 350 a year earlier. Capital-light, and compounding.
The same quarter, Amazon’s advertising services reached $17,243 million, up 24%, with worldwide paid units up 15%. Amazon disclosed that nearly 20% of shoppers who interact with a brand prompt inside Rufus, its shopping assistant, continue the conversation about that brand. That is an engagement metric, not a conversion metric, and it is worth holding the distinction — a continued conversation is not a scanned item. The consumer-facing half of this shift, including agent-driven checkout and AI-referred traffic, is covered in depth in our agentic commerce statistics analysis.
Is AI Moving the Retail Sales Line Yet?
Here is the paradox that makes retail’s low adoption rate strange rather than merely disappointing.
| Quarter | Total retail sales | E-commerce sales | E-commerce share |
|---|---|---|---|
| Q1 2025 (r) | $1,857,201M | $297,701M | 16.0% |
| Q2 2025 | $1,865,818M | $304,464M | 16.3% |
| Q3 2025 | $1,893,291M | $310,817M | 16.4% |
| Q4 2025 (r) | $1,900,653M | $317,999M | 16.7% |
| Q1 2026 (p) | $1,929,027M | $326,740M | 16.9% |
Seasonally adjusted, not adjusted for price changes. Source: U.S. Census Bureau, Quarterly Retail E-Commerce Sales, CB26-81, May 18, 2026.
E-commerce grew 9.8% year over year in Q1 2026 while total retail grew 3.9% — Axis Intelligence Research calculates a growth multiple of 2.52×, and a share gain of 0.91 points in twelve months. Strip e-commerce out and the remaining $1,602,287 million of retail sales grew 2.74%. The digital channel, where recommendation ranking, search relevance, and dynamic merchandising are already machine-driven whether or not a survey respondent calls it AI, is doing roughly two and a half times the work of the channel that is not.
That is the tension worth sitting with. The most AI-saturated part of retail is the part growing fastest, and the sector’s firm-level adoption rate is the lowest of any major services sector Census tracks. Both facts are correct because retail’s sales are concentrated in a handful of firms while retail’s businesses are not. For the broader channel picture see our e-commerce statistics and Shopify statistics datasets.
For scale: the National Retail Federation forecast 2026 U.S. retail sales of $5.6 trillion, up 4.4% over 2025, against a ten-year average of 3.6% excluding the pandemic years. NRF’s definition excludes auto dealers, gas stations, and restaurants, which is why it does not reconcile to the Census total above — different fences around the same field.
AI in Retail: United States vs Europe
Eurostat put EU-wide enterprise AI use at 19.95% in 2025 among enterprises with ten or more employees, up 6.47 percentage points year over year, from a survey of 157,000 of the EU’s 1.53 million qualifying enterprises.
| Measure | Value | Year | Source |
|---|---|---|---|
| EU enterprises using AI | 19.95% | 2025 | Eurostat, isoc_eb_ai |
| EU large enterprises (250+) | 55.03% | 2025 | Eurostat, isoc_eb_ai |
| EU medium enterprises (50–249) | 30.36% | 2025 | Eurostat, isoc_eb_ai |
| EU small enterprises (10–49) | 17% | 2025 | Eurostat, isoc_eb_ai |
| Information and communication | 62.52% | 2025 | Eurostat, isoc_eb_ain2 |
| Professional, scientific, technical | 40.43% | 2025 | Eurostat, isoc_eb_ain2 |
| All other activities | Below 25% | 2025 | Eurostat, isoc_eb_ain2 |
| Denmark (highest) | 42.03% | 2025 | Eurostat, isoc_eb_ai |
| Romania (lowest) | 5.21% | 2025 | Eurostat, isoc_eb_ai |
Eurostat’s enterprise AI statistics place every economic activity outside information and professional services below 25%, in a band running from 24.82% for real estate down to 10.79% for construction. Wholesale and retail trade sits inside that band.
A direct US-versus-EU retail comparison is not available and Axis Intelligence Research will not manufacture one. Census BTOS covers businesses of all sizes with a two-week reference window; Eurostat covers only enterprises with ten or more employees with an annual reference window and a different definition of what counts as AI. A retail-to-retail number stitched from the two would be a methodology collision wearing a percentage sign. The honest comparison is directional: on both sides of the Atlantic, retail sits in the second tier, and the leading sector runs roughly two to four times ahead of it.
Where the EU data is genuinely superior is on the question American surveys skip.
Why Are Retailers Not Adopting AI?
Eurostat asked enterprises that considered AI and declined why, and the ranking is unsentimental.
| Reason for not adopting AI | Share of enterprises that considered AI | Source |
|---|---|---|
| Lack of relevant expertise | 70.89% | Eurostat, isoc_eb_ai, 2025 |
| Lack of clarity about legal consequences | 52.52% | Eurostat, isoc_eb_ai, 2025 |
| Concerns about data protection and privacy | 48.83% | Eurostat, isoc_eb_ai, 2025 |
| Not considered useful for the enterprise | 20.68% | Eurostat, isoc_eb_ai, 2025 |
Cost does not top this list. Skills do, by a margin of eighteen points over the next reason. The number that should worry vendors is the smallest one: only 20.68% of non-adopters think AI is useless. Roughly four in five believe it would help and are not doing it anyway, which reframes the retail AI market as a staffing and legal-clarity problem rather than a product problem. Among EU enterprises not using AI, 14.21% had at least considered it, up 2.02 points on 2024.
Sarah Mitchell: Seventy-one percent citing missing expertise is the most actionable statistic in this dataset. A regional grocer does not need a foundation model. It needs one person who can tell whether the forecasting vendor’s 12% waste-reduction claim was measured against a real baseline or against the worst quarter on record. That hire is cheaper than the software and nobody is selling it.
Does AI Cost Retail Jobs?
The evidence available in mid-2026 does not support the displacement narrative, and it comes from the only U.S. instrument that asks firms directly.
Census working paper CES-26-25 reports that AI-related employment decreases occurred in 2% of firms, and that 66% of AI users rely on it solely to augment tasks. Worker-level task use ran at 23% of firms, or 41% employment-weighted. The paper’s regression analysis found a positive correlation between commercial performance and breadth of AI integration, alongside a distinct divergence on labour: functional breadth and operational investment were positively associated with employment decreases, while worker-task integration showed no significant link to headcount reduction once those were controlled for.
The nuance matters more than the headline. It is not workers using AI that correlates with job cuts — it is firms restructuring functions around it. Those are different decisions with different timelines, and 2026 data captures the early part of the second one. Our AI adoption by industry dataset tracks the cross-sector version of this question.
One caution on scope: BTOS asks about AI-attributed employment change, which relies on the respondent’s own attribution. A retailer that cut positions during a restructuring and coded the cause as “cost management” would not appear in the 2% — the figure is a floor, not a ceiling.
Methodology
Collection. Every figure in this article was retrieved from a primary source document during production on August 1, 2026, and logged in the accompanying CSV with its URL and retrieval date. No figure was drawn from model memory or from a secondary aggregator. Sources used: the U.S. Census Bureau’s Business Trends and Outlook Survey (America Counts analysis published May 26, 2026, covering collection from December 14, 2025 to May 3, 2026), Census working paper CES-26-25 (April 2026, reference period November 2025 to January 2026), Census Quarterly Retail E-Commerce Sales release CB26-81 (May 18, 2026), Eurostat’s Use of artificial intelligence in enterprises (data extracted December 2025, reference year 2025), Walmart Inc.’s Form 8-K Exhibit 99.1 for Q1 FY2027 (filed May 21, 2026), Amazon.com Inc.’s Q1 2026 earnings release (April 29, 2026), and the National Retail Federation’s 2026 annual sales forecast (March 18, 2026).
Formulas. RAAG™ v1.0 and RACD™ v1.0 are disclosed in full in their respective sections, with every input traceable to a CSV row. Both were computed and independently re-verified before publication. Derived figures — the 2.52× e-commerce growth multiple, the 0.91-point share gain, the 2.74% non-e-commerce growth rate, the 3.82-point US/EU function convergence spread — are marked axis_calculated = yes in the CSV with the formula recorded in method_note.
Where we declined to calculate. Two comparisons that readers will expect are absent by decision. We did not produce a blended global retail AI adoption rate from Census and Eurostat inputs, and we did not produce a single US-vs-EU retail adoption gap. The instruments differ on size threshold, reference window, and the definition of AI. Merging them would generate a number with no defensible meaning. The component figures are published side by side instead.
Limitations. Self-reported adoption surveys understate embedded AI: a retailer using an ERP module with machine-learning demand forecasting may not report “AI use.” All BTOS sector and expected-use figures for retail are published by Census as approximations and carry that rounding into RAAG. Census figures are firm-weighted unless noted; employment-weighted readings run substantially higher and are labelled where used. Eurostat covers only enterprises with ten or more employees, excluding the micro-retailers that make up most of the EU retail universe. Amazon’s capital expenditure is not separable into retail and cloud components from public filings, which bounds what RACD can claim. Q1 2026 Census retail estimates are preliminary and subject to revision at the August 18, 2026 release.
Independence. No vendor, retailer, or platform reviewed this analysis before publication, and no commercial relationship influenced any figure or conclusion.
About This Dataset
This dataset covers U.S. and EU retail artificial intelligence adoption rates by sector and firm size, AI deployment by business function, retail capital expenditure intensity, quarterly U.S. retail and e-commerce sales from Q1 2025 to Q1 2026, barriers to adoption, and AI-attributed employment effects, for the 2025–2026 period. The RAAG™ and RACD™ metrics are original calculations by Axis Intelligence Research.
License. Creative Commons Attribution 4.0 International (CC BY 4.0). Reuse and redistribution permitted with attribution to Axis Intelligence Research and a link to this page.
Suggested citation
- APA: Axis Intelligence Research. (2026, August 1). AI in retail statistics 2026: Adoption rates, spending, and the gap nobody is pricing in. Axis Intelligence. https://axis-intelligence.com/ai-in-retail-statistics/
- MLA: Axis Intelligence Research. “AI in Retail Statistics 2026: Adoption Rates, Spending, and the Gap Nobody Is Pricing In.” Axis Intelligence, 1 Aug. 2026, axis-intelligence.com/ai-in-retail-statistics/.
- Chicago: Axis Intelligence Research. “AI in Retail Statistics 2026: Adoption Rates, Spending, and the Gap Nobody Is Pricing In.” Axis Intelligence, August 1, 2026. https://axis-intelligence.com/ai-in-retail-statistics/.
Frequently Asked Questions
What percentage of retailers use AI in 2026?
Around 14% of U.S. retail trade businesses reported using AI in any business function as of the Census Bureau’s Business Trends and Outlook Survey collection period ending May 3, 2026. The national all-sector rate for the same period was 19.8%.
Is retail ahead of or behind other industries in AI adoption?
Behind. Retail Trade’s 14% sits 5.8 percentage points below the 19.8% national rate, 2.84 times below the Information sector’s 39.7%, and 2.42 times below Finance and Insurance’s 33.9%, per Census Bureau data as of May 3, 2026.
What is the Retail AI Adoption Gap?
RAAG™ is an Axis Intelligence Research metric measuring retail’s distance from the U.S. all-sector AI adoption benchmark. Its baseline reading is 5.8 percentage points (29.3% in relative terms) as of May 3, 2026, with a 2.8-point Intent Shortfall indicating retail’s six-month forward expectation still trails the current all-sector rate.
What do retailers use AI for most?
Marketing and sales. Among EU retail trade enterprises using AI, 48.18% deployed it for marketing or sales in 2025, per Eurostat. Among U.S. AI-adopting firms across all sectors, 52% used it in Sales and Marketing, per Census working paper CES-26-25. Logistics ranked last across EU AI users at 6.08%.
How much are large retailers spending on AI?
Public filings do not break out AI as a line item, but capital expenditure intensity captures the divergence. Amazon spent 24.35% of quarterly net sales on property and equipment in Q1 2026 — attributing a $59.3 billion year-over-year increase primarily to AI — while Walmart spent 3.80% in the quarter ended April 30, 2026. Axis Intelligence Research calculates a Retail AI Capital Divergence Ratio of 6.40×.
Why are most retailers not using AI?
Skills, not cost. Among EU enterprises that considered AI and declined, 70.89% cited lack of relevant expertise, 52.52% cited unclear legal consequences, and 48.83% cited data protection concerns. Only 20.68% said AI was not useful for their business, per Eurostat 2025 data.
Is AI reducing retail employment?
Not yet, on the available evidence. Census working paper CES-26-25 found AI-related employment decreases in 2% of U.S. firms, with 66% of AI users deploying it solely to augment tasks. The paper did find that functional breadth and operational investment correlate positively with employment decreases, while worker-level task use does not.
How much of U.S. retail is e-commerce in 2026?
E-commerce accounted for 16.9% of total U.S. retail sales in Q1 2026 — $326.7 billion of $1,929.0 billion, seasonally adjusted — up from 16.0% a year earlier, per Census release CB26-81. E-commerce grew 9.8% year over year against 3.9% for total retail, a 2.52× growth multiple calculated by Axis Intelligence Research.
Does AI adoption vary by retailer size?
Substantially. Census BTOS put AI use at 37% for firms with 250 or more employees and 32% for firms with 100 to 249, against under 20% for firms with four or fewer, as of May 3, 2026. Eurostat found the same pattern in the EU: 55.03% of large enterprises versus 17% of small ones in 2025.
Where can I download the AI in retail dataset?
The full fact table ships as a CSV under CC BY 4.0, with source organisation, source document, URL, retrieval date, and calculation notes on every row. Every figure in this article appears as a CSV row at the same value and as-of date.