Research by Imagine
Enterprise AI diffusion report
The Enterprise AI diffusion report is Imagine's research on how large companies deploy AI: who owns it, whether they built or bought their platform, which models run where, how many employees use it, and what return they have reported.
Verified See methodology
How enterprises have deployed AI
By sector
Financial services
Bank of America's Erica does work that would otherwise take 10,000 to 12,000 staff.
Healthcare
99% of UnitedHealth Group's AI applications are administrative, not clinical.
Industrials
Uber used up its 2026 budget for AI coding tools in about 4 months.
Retail and consumer
Online shoppers who use Lowe's Mylow assistant convert at 3 times the rate of others.
Food, beverage and household
Walmart's AI digital twins cut its refrigeration maintenance spend 19%.
Technology
Amazon credits its Rufus shopping assistant with nearly $12 billion of incremental annualized sales.
Media and telecom
Disney dropped a planned $1 billion investment in OpenAI when OpenAI shut down Sora.
Oil and gas
Returns are often counted in barrels: about 2,000 a day more at BP's Atlantis hub.
Utilities
NextEra Energy sells its grid AI to other utilities and keeps all of the intellectual property.
Materials
At Linde, the chief human resources officer oversees AI rather than IT.
Method
What the report measures
- Ownership
- Who owns AI on the org chart, and how close that person sits to the chief executive.
- Platform
- Whether the company built its own AI platform, bought one, or runs a mix of both.
- Models
- Which model providers the company uses, and whether it can switch between them.
- Hosting
- Where inference runs, and whether company data leaves the company's own environment.
- Reach
- How many employees can use AI tools, and how many use them routinely.
- Stage
- The return the company has reported from AI, from experimenting to embedded.
Stages of diffusion
From experimenting to embedded
1Experimenting
Experimenting means the company has reported neither a business return it credits to AI nor a company-wide AI rollout with a user count. Pilots, partnerships, plans and spending alone keep a company at Experimenting.
Maersk, Abbott, AIG, Air Products, Allstate, American, Apple, AstraZeneca, AutoZone, Berkshire, Best Buy, BHP, BlackRock, Boeing, Burlington, Capital One, Charter, Chipotle, Chubb, Comcast, ConocoPhillips, Costco, CVS Health, Dell, Deloitte, Delta, Disney, Dollar General, Dominion, Domino's, Dow, Duke Energy, Duolingo, Lilly, Enbridge, Exelon, ExxonMobil, Ford, Freeport, GE Aerospace, GM, HCA, Honeywell, Humana, Iberdrola, Deere, J&J, Kaiser, Linde, Marriott, McDonald's, Medtronic, MetLife, Mondelez, Morgan Stanley, National Grid, Netflix, NextEra, Nike, Novartis, Nucor, Oracle, PepsiCo, Pfizer, PG&E, P&G, Progressive, Rio Tinto, Ross, Samsung, Shell, Sherwin-Williams, Siemens, Southern, Southwest, Spotify, Starbucks, T-Mobile, Tesla, Cigna, Home Depot, TotalEnergies, Toyota, Tyson, U.S. Bank, Union Pacific, United, UnitedHealth, UPS, WBD
2Piloting
Piloting means the company reports a business return it credits to AI in at least one use case, or a company-wide AI rollout with a user count and a named owner.
Accenture, Amazon, Amex, AT&T, BNY, BP, Caterpillar, Chevron, Citi, Coca-Cola, Colgate, Elevance, Goldman, HSBC, IBM, Intel, Intuit, Kroger, L'Oréal, Lowe's, LVMH, Mastercard, McKinsey, Mercedes-Benz, Merck, Meta, Moderna, Nestlé, NVIDIA, Occidental, PwC, SAP, Shopify, Target, Travelers, Uber, Unilever, Vale, Verizon, Visa, Vodafone, Walmart, Wells Fargo
3Scaling
Scaling means the company reports returns it credits to AI in several functions, or a firm-level AI figure, and runs a company-wide AI platform with a named owner.
Google, AMD, BofA, DBS, Equinor, FedEx, JPMorgan, Lemonade, Microsoft, Salesforce, Aramco, ServiceNow
4Embedded
Embedded means the company reports AI's effect on the whole business, in cost, headcount, revenue or margin, and most employees use AI every week in redesigned workflows.
Unknown
Unknown means the company has said too little about its AI work in the last 18 months to place it: no AI return, rollout, program, pilot or tool on the public record in that time.
Methodology
How a profile is built
- Sources. Facts come from the company's own annual reports, shareholder letters, investor days, earnings calls and announcements. News coverage and executive interviews fill gaps, and a claim with only one secondary source is marked unconfirmed.
- Returns. A return counts only when the company reported it, credited it to AI alone and realized it: revenue, cost, headcount, losses avoided or output per employee, with its date. Usage, time saved, volume measures, pilots, targets and returns shared with other technology do not count.
- Stages. A company's stage follows from the returns and rollouts it has reported in the last 18 months, not from its spending or plans.
- Confidence. Every figure is marked confirmed, estimated or unconfirmed, and an estimate says what it is derived from.
- Dates. Each profile shows the day its facts were last checked against their sources.
Data confidence
- Confirmed
- Stated in a primary or reputable public source.
- Estimated
- Derived or approximated from public reporting.
- Unconfirmed
- Reported, but not yet corroborated.
Questions
About the report
- What is the Enterprise AI diffusion report?
- The Enterprise AI diffusion report is Imagine's research on how large companies deploy AI: who owns it, whether they built or bought their platform, which models run where, how many employees use it, and what return they have reported.
- How does the Enterprise AI diffusion report decide a company's stage?
- Imagine places each company on one of four stages by the returns it has publicly reported from AI, not by its spending or plans. One return credited to AI, or a company-wide rollout with a user count and a named owner, makes a company Piloting. Returns in several functions, or a firm-level AI figure, make it Scaling. A reported effect on the whole business, with most employees using AI every week, makes it Embedded. Every other company is Experimenting, including any that see returns but have not reported them. Only evidence reported in the last 18 months counts, and a company that has said too little about its AI work in that time is marked Unknown.
- Where does the data in the Enterprise AI diffusion report come from?
- Every profile in the Enterprise AI diffusion report is built from public sources: annual reports and shareholder letters, investor day presentations, earnings calls, company announcements, executive interviews and news reporting. Each figure is marked confirmed, estimated or unconfirmed, and each profile shows the date its facts were last checked.
- Which companies are in the Enterprise AI diffusion report?
- The Enterprise AI diffusion report profiles large companies in financial services, healthcare, industrials, retail and consumer, food, beverage and household, technology, media and telecom, oil and gas, utilities, and materials, among them JPMorgan Chase, UnitedHealth Group, ExxonMobil, and Klarna. Companies are grouped by their GICS sector and industry group, and each is measured on the same six dimensions: ownership, platform, models, hosting, reach and stage.
- How can I suggest a company or report a correction?
- Email data@imagine.ai with the company name or the correction and a source. Imagine checks every correction against its source before changing a profile, and the profile's verification date moves when it does.