Generative AI Statistics 2026: Adoption, Usage, Spending & Growth
OpenAI says ChatGPT now reaches 1.2 billion people every week, Google says its Gemini app passed 1 billion monthly users in August 2026, and Gartner forecasts $2.67 trillion of worldwide AI spending this year. Yet the best enterprise survey of 2026 shows a gap: nearly nine in ten organizations use AI, while only 37% report any profit impact (McKinsey). This guide collects the strongest generative AI statistics for 2026, dates every figure, separates observed data from forecasts, and explains why the numbers disagree.
TL;DR
Consumer use is at internet scale. ChatGPT reported 1.2 billion weekly users on September 29, 2026, and the Gemini app passed 1 billion monthly users on August 11, 2026. The two metrics differ and cannot be added together.
US workplace use has passed the halfway line. Gallup found 52% of US employees used AI at work in May 2026, and 30% used it a few times a week or more. Microsoft measured 18.8% of working-age people worldwide using generative AI in Q2 2026.
Scaling is up, profit impact is flat. McKinsey’s 2026 survey found 44% of organizations scaling AI across the enterprise, yet 37% report any EBIT impact, unchanged from 2025.
Money flows to infrastructure and a few giant companies. Gartner forecasts $2.67 trillion of AI spending in 2026, and Crunchbase says OpenAI and Anthropic took 43% of global startup funding in the first half.
Governance lags adoption. IBM found shadow AI in 43% of the breach incidents it studied, up from 20%, and 68% of breached organizations lacked AI governance policies.
Generative AI is mainstream in 2026 but uneven. OpenAI reports 1.2 billion weekly ChatGPT users, Gallup finds 52% of US employees use AI at work, and McKinsey finds nearly nine in ten organizations use AI in at least one function. Only 37% report enterprise-level profit impact, so adoption is outrunning measured returns.
Which statement best describes your organization’s use of generative AI in 2026?
0%We are not using generative AI
0%We are evaluating it, but have not started a pilot
0%We are piloting it in one or more teams
0%We use it in production for a few workflows or teams
Table of Contents
Generative AI Statistics 2026 at a Glance
The table lists the most useful verified figures. Each row has its own period, so compare rows with care, and note that forecasts are labeled. For a wider view that includes non-generative AI, see our Artificial Intelligence Statistics 2026 guide.
Metric | Statistic | Period | Source |
ChatGPT weekly users | 1.2 billion | Sept 29, 2026 | |
Gemini app monthly users | 1 billion+ | Aug 11, 2026 | |
US adults using AI chatbots | 49% | Survey, Feb 2026 | |
Working-age people using generative AI, worldwide | 18.8% | Q2 2026 | |
US employees using AI at work | 52% | May 2026 | |
Organizations regularly using AI in one or more functions | Nearly 9 in 10 | May–June 2026 | |
Organizations scaling AI across the enterprise | 44% | May–June 2026 | McKinsey |
Large companies scaling AI agents | 40% (up from 27%) | May–June 2026 | McKinsey |
Respondents reporting any EBIT impact from AI | 37% (flat) | May–June 2026 | McKinsey |
US businesses using AI in a business function | 19.8% | As of May 3, 2026 | |
EU enterprises (10+ staff) using AI | 20.0% | 2025 | |
Enterprise spending on generative AI | $37 billion | 2025 | |
Global private AI investment | $344.7 billion (generative AI: $170.9 billion) | 2025 | |
Global venture funding | $510 billion; 43% to OpenAI and Anthropic | H1 2026 | |
Worldwide AI spending (forecast) | $2.67 trillion, +49.5% | 2026 forecast | Gartner |
Generative AI model spending (forecast) | $28.3 billion, +117% | 2026 forecast | Gartner |
Developers using AI tools | 79% | 2025 survey | |
Wage premium for AI skills | 62% | 2026 report | |
Breach incidents involving shadow AI | 43% (from 20%) | Mar 2025–Feb 2026 |
Five patterns stand out.
Use is ahead of payoff. In McKinsey’s 2026 survey, 80% of respondents said AI improved their own productivity, yet only 37% reported any enterprise-level EBIT impact.
Employees move faster than employers. Gallup finds 52% of US employees use AI at work, while 47% say their organization has integrated AI tools. IBM ties the gap to shadow AI.
The yardstick decides the headline. Census counts 19.8% of US businesses using AI, while McKinsey’s survey, tilted toward larger organizations, shows nearly nine in ten. Both can be right, as the adoption section explains.
Capital is concentrated. OpenAI and Anthropic took 43% of global startup funding in the first half of 2026, and Gartner puts infrastructure as the largest share of AI spending.
Agents are a large-company story so far. McKinsey found 40% of large organizations scaling AI agents, against 22% of smaller ones.
How Many People Use Generative AI in 2026?
Short answer: it depends on who is counted. Microsoft measured 18.8% of the world’s working-age population (ages 15–64) using generative AI in Q2 2026. Pew found 49% of US adults use AI chatbots, and a study cited by Stanford’s AI Index put US adult use of generative AI at 56%.
Platform figures are larger but use different definitions. OpenAI reports 1.2 billion weekly ChatGPT users, up from the 900 million that trade press reported in February 2026. Google reports monthly users for the Gemini app, which grew from 400 million in May 2025 to more than 1 billion on August 11, 2026, and it has not explained how it defines a monthly active user (SiliconANGLE). Weekly and monthly counts overlap, so they should not be added.
Pew’s February 2026 survey of 5,119 US adults found ChatGPT used by 44%, Gemini by 24%, Copilot by 17%, Meta AI by 14% and Claude by 6%. Among adults under 30, 61% use ChatGPT. Pew’s ChatGPT series rose from 18% in 2023 to 44% in 2026, but the 2026 question also named other chatbots, so the trend needs a caution.
Microsoft’s measure is telemetry-based and adjusted for device mix, internet access and population. It shows 28.8% in the Global North and 16.2% in the Global South in Q2 2026. In Q1 2026, the UAE led at 70.1% and the US ranked 21st at 31.3%, well below survey-based US figures.
That gap explains much of the confusion. Surveys ask people whether they use AI, while telemetry counts observed use on one company’s products and then adjusts it. Stanford also reports that generative AI reached about 53% adoption within three years of mass-market launch, faster than the personal computer or the internet at the same stage, and that four in five university students use it.
Most consumer tools are free or cheap. Stanford’s summary of a Brynjolfsson et al. study estimates US consumer surplus from generative AI at $172 billion a year by early 2026, up from $112 billion, with the median value per user rising from $3.40 to $11.40. That is a value estimate, not revenue.
Generative AI Adoption Statistics
Short answer: about nine in ten respondents in a large global survey say their organization uses AI regularly, while government counts of all businesses find about one in five. The definition drives the number.
McKinsey’s 2026 survey (1,719 respondents in 97 nations, May 4–June 8, 2026) found nearly nine in ten report regular AI use in at least one function. It also found 44% scaling AI across the enterprise, up from 38%, and 56% using AI in three or more functions, up from 51%. A year earlier, McKinsey’s data showed 88% using AI and 79% using generative AI regularly in at least one function, up from 78% and 71% in 2024, according to Stanford.
Measure | Figure | Who is counted | Source |
Regular AI use in one or more functions | Nearly 9 in 10 | Survey respondents in 97 nations; 36% from organizations above $1 billion revenue | McKinsey, 2026 |
Regular generative AI use in one or more functions | 79% (71% in 2024) | Same survey series, 2025 | McKinsey via Stanford |
AI use in a business function | 19.8% | US businesses of all sizes, as of May 3, 2026 | Census Bureau BTOS |
Use of AI technologies | 20.0% | EU enterprises with 10+ employees, 2025 | Eurostat |
Organization has integrated AI tools | 47% | US employees, May 2026 | Gallup |
Pilots are not production. In McKinsey’s 2025 data, larger companies were more likely to report scaled AI programs, and scaled AI agent use was in the single digits in almost every function, per Stanford. By 2026, 40% of large organizations report scaling agents in at least one function.
The Census Bureau’s own AI supplement found 18% of firms using AI in a business function between November 2025 and January 2026. Its monthly survey stayed between 17% and 20% from December 2025 to May 2026, as reported by Digital Watch.
Enterprise and Workplace Generative AI Usage
Short answer: just over half of US employees use AI at work, but fewer than a third use it often. In Gallup’s survey of 22,573 US employees (May 6–20, 2026), 52% used AI at least a few times a year, 30% a few times a week or more, and 15% daily. In Q1 2026 the figures were 50%, 28% and 13%, against 21% reporting any use in Q2 2023.
Employers are catching up more slowly. Gallup found 47% of employees say their organization has integrated AI tools, up from 41% a quarter earlier, while about one in five are unsure. Daily use is concentrated: 42% in technology, 27% in finance and 22% in professional services, versus 9% to 15% in the other seven industries (Gallup AI indicator).
Employees most often use AI for writing, research and problem-solving; 18% use it for data science or analytics and 17% for slide decks. Frequent users are nearly three times as likely to use it for coding help.
McKinsey’s 2026 survey adds cost and strain. About 28% of respondents say their organization spends more than 10% of its ICT budget on AI, and 20% say AI operating costs, including tokens, constrain use. Among mid-level managers and individual contributors, 47% report at least one AI-related strain, against 31% of executives and senior managers.
Unofficial use fills the gap between employee habits and employer programs. IBM’s 2026 breach study found shadow AI in 43% of incidents, up from 20%, as the governance section details.
[CHART OPPORTUNITY: Line chart of the share of US employees using AI at work (any use, %): 21% in Q2 2023, 45% in Q3 2025, 46% in Q4 2025, 50% in Q1 2026 and 52% in Q2 2026. Source set: Gallup Workforce Panel.]
Most Common Generative AI Use Cases
Short answer: consumers mostly ask for advice, information and writing, while enterprises pay mainly for coding and general-purpose assistants.
An NBER working paper by OpenAI and Harvard researchers analyzed ChatGPT consumer conversations through mid-2025. Practical guidance, seeking information and writing made up nearly 80% of conversations. Non-work use grew from 53% to more than 70%, writing was about 40% of work messages, and computer programming was just 4.2% of all messages. The data cover consumer plans only and were classified automatically.
Anthropic’s Economic Index, which tracks Claude, shows a different mix. Computer and mathematical tasks accounted for close to 40% of Claude.ai activity through 2025, and educational tasks grew from 9% to about 14%. In November 2025, 52% of conversations were augmentation, where users work with the model, against 45% automation, per Stanford. Usage then diversified: the ten most common tasks fell from 24% of Claude.ai conversations in November 2025 to 19% in February 2026 (Anthropic). These are one vendor’s data and reflect its own users.
Among EU enterprises in 2025, 11.8% used AI to analyze written language, 9.5% to generate images, video or audio, and 8.8% to generate written or spoken language (Eurostat).
In spending terms, Menlo Ventures estimates that $19 billion of 2025 enterprise generative AI spend went to applications: $8.4 billion to horizontal assistants and copilots, $7.3 billion to department-specific tools (coding was $4.2 billion of that) and $3.5 billion to industry-specific tools. The estimate combines a bottom-up model with a survey of about 495 US enterprise decision-makers. For classroom use, see our AI in Education guide.
Generative AI Adoption by Business Function
Function-level data mostly measures AI broadly, not only generative AI, so read it as a guide to where generative tools land first.
Software engineering and IT
In McKinsey’s 2025 data, reported by Stanford, AI use was highest in software engineering (58%) and IT (56%) within the technology sector. In 2026, respondents most often reported scaling AI agents in IT, knowledge management and software engineering, and about two in ten were scaling software coding agents (31% at larger enterprises).
Marketing and sales
McKinsey’s 2026 respondents most often attributed revenue gains to marketing and sales. In the Census AI supplement, 52% of AI-using firms applied AI in sales and marketing, the most common function. A study of marketing teams using multimodal ad creation found 50% more output per worker (Ju and Aral, 2025, via Stanford).
Customer service and operations
McKinsey’s 2026 respondents most often reported cost reductions in supply chain management, service operations and manufacturing. A study of customer support agents found 14% to 15% more issues resolved per hour, and 30% to 35% for less experienced agents (Brynjolfsson et al., 2025, via Stanford). Service operations also ranked among the highest for expected headcount reductions in McKinsey’s 2025 survey.
Knowledge work, finance and legal
Stanford’s summary of McKinsey data put AI use highest in knowledge management for business, legal and professional services (58%). Uptake stayed low in strategy and corporate finance and in risk and compliance across most sectors, with financial services the exception for risk and compliance. Census data show 33.9% of finance and insurance businesses using AI, against 19.8% nationally.
Generative AI Adoption by Industry, Company Size and Region
Short answer: adoption is highest in technology, information and finance, in larger firms, and in a handful of wealthy or digitally focused economies.
Cut | Group | Statistic | Source |
Company size (US) | 250+ employees | 37% using AI | Census BTOS, May 3, 2026 |
Company size (US) | 100–249 employees | 32% | Census BTOS |
Company size (US) | 4 or fewer employees | Under 20% | Census BTOS |
Industry (US) | Information | 39.7% | Census BTOS |
Industry (US) | Finance and insurance | 33.9% | Census BTOS |
All US businesses | Average | 19.8% | Census BTOS |
Region (working-age) | Global North vs Global South | 28.8% vs 16.2% | Microsoft, Q2 2026 |
EU enterprises | Denmark vs Romania | 42.0% vs 5.2% | Eurostat, 2025 |
Size gaps show up in agents too. McKinsey found 54% of organizations above $1 billion in revenue scaling AI across the enterprise, versus one-third of smaller ones, and 40% versus 22% scaling agents.
By industry, McKinsey found agents most widely reported in technology and in media and telecom, with consumer goods and retail using them in marketing and sales and advanced manufacturing in supply chain and production. The share of respondents who skipped buying software because coding agents could build it in-house was highest in technology and healthcare.
Regional gaps are wide. In the EU, Denmark (42.0%), Finland (37.8%) and Sweden (35.0%) sit far above Poland (8.4%) and Romania (5.2%). Stanford’s summary of McKinsey’s 2025 survey put organizational AI use at 91% in Europe and 90% in North America, with China and Europe posting the biggest gains, 13 and 11 percentage points. Survey samples are not globally representative, so treat regional rankings as indicative.
Generative AI Market Size and Growth
Short answer: there is no single market size. Gartner’s September 2026 forecast puts worldwide AI spending at $2.67 trillion in 2026, but only $28.3 billion of that is generative AI models, up 117% from $13.0 billion in 2025. The rest is infrastructure, software, services and other categories.
Category (US$ billions) | 2025 | 2026 (forecast) | 2027 (forecast) |
AI infrastructure | 981.9 | 1,484.4 | 1,977.7 |
AI services | 434.0 | 576.5 | 745.7 |
AI software | 288.2 | 461.6 | 656.4 |
AI agents and assistants | 16.5 | 29.2 | 65.5 |
Generative AI models | 13.0 | 28.3 | 51.6 |
AI cybersecurity | 25.9 | 51.3 | 86.0 |
Total AI spending | 1,786.7 | 2,670.5 | 3,637.3 |
Source: Gartner, September 2026. Categories shown are a subset of the full table, so rows do not sum to the total.
Scope changes the answer. In a July 2026 release, Gartner put end-user spending on AI platforms and models at $64 billion for 2026, up 63.4% from $39 billion, with foundation generative AI models at $23.4 billion (+104.2%). That uses a narrower definition and an earlier forecast date than the $28.3 billion above.
Bottom-up estimates differ again. Menlo Ventures, which surveyed US enterprise decision-makers, counted $37 billion of enterprise generative AI spend in 2025, up from $11.5 billion in 2024 (3.2×), including $19 billion on applications. Stanford’s AI Index, citing Epoch AI, shows annualized revenue of roughly $25 billion at OpenAI and $19 billion at Anthropic in early 2026.
Forecasts also move. Gartner’s 2026 worldwide AI spending forecast rose from $2.52 trillion in January to $2.59 trillion in May and $2.67 trillion in September, so check the publication date before quoting any forecast.
[CHART OPPORTUNITY: Stacked bar chart of Gartner’s AI spending by category (US$ billions) for 2025 (estimate) and 2026–2027 (forecast), excluding the total row. Source: Gartner, September 2026.]
Generative AI Spending and Investment
Enterprise spending
Menlo Ventures reports $37 billion of enterprise generative AI spending in 2025, about 3.2 times 2024. In McKinsey’s 2026 survey, 28% of respondents spend more than 10% of their ICT budget on AI, 60% expect to raise AI investment next year, and 32% decided against buying a software product because coding agents could build it. Gartner places generative AI in the Trough of Disillusionment in 2026 and says enterprises are mostly buying simpler embedded AI features from incumbent software vendors.
Infrastructure and capital spending
Infrastructure is the largest block: about 56% of Gartner’s $2.67 trillion total (a calculation from Gartner’s table). Gartner projects AI-optimized IaaS at $42.3 billion in 2026 (+96%), with inference spending ($23.3 billion) overtaking training ($19.0 billion). IDC projects $497 billion of AI infrastructure spending in 2026 (+56%); the totals differ because Gartner’s category is broader. Alphabet raised its 2026 capital spending forecast to $195–205 billion, and frontier-lab compute spend in 2025 reached $16.3 billion at OpenAI and $6.8 billion at Anthropic (Epoch AI, via Stanford).
Venture and private investment
Stanford’s AI Index, using Quid data, counts $344.7 billion of global private AI investment in 2025 (+127.5%), including $170.9 billion in generative AI companies, and $581.7 billion of total corporate AI investment once mergers and acquisitions and other categories are included. The US drew $285.9 billion and had 28 funding events above $1 billion.
Crunchbase counted a record $510 billion of global venture funding in the first half of 2026, with $217 billion (43%) going to OpenAI and Anthropic and more than 70% of Q2 capital going to AI-focused companies. PitchBook-NVCA put US venture at $412.7 billion, 86% of it to AI, as reported by AI Weekly.
Do not add these figures together. Venture funding, corporate capital spending, M&A and enterprise purchases measure different flows, and providers define an “AI company” differently.
Generative AI ROI, Productivity and Business Value
Short answer: individual productivity gains are widely reported and measurable in narrow tasks, but enterprise-level profit impact is not yet widespread.
In McKinsey’s 2026 survey, 80% of respondents said AI improved their own productivity and about half said it helped them make better decisions, yet 37% reported any EBIT impact, unchanged from 2025. About 6% qualify as high performers, attributing at least 5% of EBIT to AI. Nearly three-quarters of them have fundamentally redesigned workflows, against one-quarter of other respondents, and they are twice as likely to have defined processes for measuring impact.
Deloitte’s survey of 3,235 leaders in 24 countries, fielded in August and September 2025, found 66% reporting productivity or efficiency gains and 34% using AI to reimagine the business.
Study | Setting | Finding |
Brynjolfsson et al., 2025 | Customer support agents | 14%–15% more issues resolved per hour |
Cui et al., 2025 | Developers using GitHub Copilot | 26% more completed pull requests |
Ju and Aral, 2025 | Marketing teams using AI ad creation | 50% more output per worker |
Becker et al. (METR), 2025 | Experienced open-source developers | 19% slower with AI; a later METR study could not replicate this |
Aldasoro et al., 2026 | About 12,000 European firms | 4% higher labor productivity |
Yotzov et al., 2026 | 6,000 executives in four countries | High adoption, minimal realized productivity gains |
Studies as summarized in Stanford’s AI Index 2026.
Gains are largest in structured, measurable work, and less experienced workers often benefit most. METR’s 2025 result shows perceived and measured speed can diverge; METR later said developers’ reluctance to work without AI prevented a replication, and that late-2025 developers were probably sped up. At the macro level, US productivity growth reached 2.7% in 2025 against a 1.4% average over the prior decade, which Stanford says may reflect an early J-curve, but the evidence remains early and mixed.
PwC found productivity growth 40% higher at companies most exposed to AI than at the least exposed, a correlation that does not prove AI caused it. For return benchmarks by use case, see our AI Agent ROI Benchmarks analysis.
Developer and AI Coding Statistics
Short answer: most developers now use AI tools, and agent use roughly doubled in a year, but enthusiasm and trust have cooled. Stack Overflow reports that current AI tool use among its respondents rose from 44% in 2023 to 62% in 2024 and 79% in 2025, with almost half using them daily (47.1%). Its 2026 survey results had not been published when this article was written.
Agents moved fastest. Stack Overflow found 31% of 2025 respondents using AI agents, and 59% in a smaller April 2026 pulse survey. Because the two samples differ, treat the jump as directional. Favorable sentiment toward AI tools among all respondents fell from 72.0% in 2024 to 59.7% in 2025 while negative sentiment rose from 6.4% to 20.4%.
In 2025 work-tool data, Cursor was used by 17.9% of respondents and Claude Code by 9.7%, alongside rather than replacing existing tools. In McKinsey’s 2026 survey, about two in ten organizations were scaling software coding agents (31% at larger enterprises), and 32% had decided against buying software because coding agents could build it. Menlo Ventures estimates $4.2 billion of 2025 enterprise spend went to coding tools.
Output evidence is mixed. A study of GitHub Copilot users found 26% more completed pull requests (Cui et al., 2025), while METR’s 2025 trial found experienced developers 19% slower, a result METR could not replicate later. Microsoft’s Q1 2026 diffusion report, as covered by Redmondmag, found US developer employment about 4% higher in March 2026 than a year earlier. Stanford, by contrast, reports employment for software developers aged 22 to 25 down nearly 20% from its 2022 peak by September 2025. Our Agent Skills guide explains how coding agents are being extended.
Generative AI Workforce and Jobs Statistics
Short answer: large job losses have not shown up in aggregate data, but expectations of cuts are rising and early-career workers in exposed jobs look most affected.
In McKinsey’s 2026 survey, 39% of respondents expect AI to reduce their organization’s total employment in the next year, up from 32% last year, and 43% expect no change. Yet only 14% of respondents from AI-using organizations reported an AI-related decline in workforce size over the past year, less than half the 32% who had expected one. Just 13% said AI makes them anxious about their career.
Stanford’s analysis finds early-career workers in the most AI-exposed occupations had about 16% lower employment than the least exposed after controls, while the unemployment rate rose more for the least exposed group. It concludes the evidence does not point to broad, uniform displacement. A survey of 844 tasks found 46.1% of workers want AI to take over those tasks.
Indicator | Statistic | Source |
Wage premium for AI skills | 62% (57% a year earlier) | PwC, 2026 |
Growth in jobs requiring AI skills | 69%, vs 9% for all jobs | PwC, 2026 |
US job postings requiring AI skills | 2.6% in 2025 (Singapore 4.69%) | Lightcast via Stanford |
Workers wanting AI to take over a task | 46.1% of tasks surveyed | Shao et al., 2026, via Stanford |
PwC’s 2026 Barometer analyzed more than one billion job ads in 27 countries. Its wage premium is an association in job advertisements, not proof that AI skills cause higher pay. For why firms need these skills internally, see our AI Center of Excellence guide.
Security, Governance and Adoption Barriers
Security and governance
IBM’s 2026 Cost of a Data Breach Report, based on 602 breached organizations between March 2025 and February 2026, put the global average breach cost at $4.99 million, up 12%. Shadow AI, meaning unapproved AI tools, was involved in 43% of incidents, up from 20%, and one in four malicious breaches was AI-enabled, at about $6 million each. Reporting on the study adds that shadow AI breaches averaged $5.39 million, 68% of breached organizations lacked AI governance policies, and 92% of those with an AI-related breach lacked proper AI access controls.
McKinsey found AI high performers are much more likely to work on mitigating AI-related risks such as exploited vulnerabilities and unauthorized actions. Our guides on generative AI security and AI security governance cover controls in depth.
Adoption barriers
Unclear value for the individual. Gallup research cited by The Register found lack of utility is the most common barrier to individual AI use.
Operating costs. About 20% of McKinsey’s respondents say AI operating costs, including tokens, constrain use.
Workflow redesign. Nearly three-quarters of high performers redesigned workflows; only one-quarter of others did.
Disinterest. Pew’s 2026 survey found 51% of US adults do not use chatbots, and 60% of them cited disinterest rather than access or skill (The Next Web).
Trust and learning risk. Stanford cites research that heavy reliance on AI while learning showed no speed gain and a “learning penalty.”
Generative AI Growth Forecasts
All figures below are forecasts, not observed results, and each firm defines its market differently.
Forecast | Figure | Publisher, date |
Worldwide AI spending, 2027 | $3.64 trillion | Gartner, Sept 2026 |
Generative AI model spending, 2027 | $51.6 billion | Gartner, Sept 2026 |
AI agents and assistants spending, 2027 | $65.5 billion | Gartner, Sept 2026 |
AI services opportunity, 2030 | $1.2 trillion | Gartner, Sept 2026 |
AI-optimized IaaS, 2027 | $66 billion | Gartner, Aug 2026 |
AI infrastructure, 2029 | $1.08 trillion | IDC, 2026 |
Total AI spending, 2029 | About $1.3 trillion (31.9% growth rate) | IDC, Aug 2025 |
Asia/Pacific generative AI spending, 2029 | About $175 billion (68.2% CAGR) | IDC, as reported, Apr 2026 |
The IDC and Gartner totals differ because IDC’s older total uses a different scope and date than Gartner’s September 2026 view, so they should not be compared line by line. Plans also point up: 60% of McKinsey’s 2026 respondents expect to raise AI investment next year, though that is an expectation, not a spending forecast.
What the Numbers Mean for Businesses in 2026
This section separates evidence from recommendation.
Where adoption looks mature. Individual and chatbot use is mature: McKinsey reports 47% scaling chatbots enterprise-wide, and Gallup finds a majority of US employees using AI. Where it is experimental. Agents are scaling mainly at large firms (40% versus 22%), and Stanford found scaled agent use in the single digits in most functions in 2025.
Strongest evidence of value. Narrow, measurable tasks: customer support, coding and marketing production show gains in named studies. Weakest evidence. Enterprise-wide profit impact, which only 37% report, and which rests on self-reported surveys. Treat vendor and survey claims of ROI as directional.
Where spending is growing. Infrastructure leads, and Gartner expects enterprises to buy mostly embedded features from incumbent vendors. McKinsey’s finding that 32% skipped a software purchase suggests build-versus-buy decisions are being reopened, while 20% report cost constraints.
What to measure before scaling. Baseline cycle time and quality for a named workflow, cost per task including tokens, active use by role, and the share of use that is unsanctioned. Governance. IBM’s numbers tie ungoverned use to higher breach costs; see our AI risk management and NIST AI RMF guides. People. Early-career roles look most exposed, and PwC’s data show a growing premium for AI skills, so training and entry-level pathways deserve budget.
Priorities by role. Executives: fund a few redesigned workflows rather than many pilots. Founders and SMB leaders: Census data show use is lower in small firms, so a single well-measured use case is a reasonable start; see our enterprise AI strategy guide. Technology leaders: inventory shadow AI and set token budgets. Marketing and operations leaders: start where output is easy to measure.
FAQ
How many people use generative AI in 2026?
OpenAI reported 1.2 billion weekly ChatGPT users on September 29, 2026, and Google said the Gemini app passed 1 billion monthly users on August 11, 2026. Microsoft measured 18.8% of working-age people worldwide using generative AI in Q2 2026. Pew found 49% of US adults use AI chatbots in February 2026. These figures use different definitions and cannot be added together.
What percentage of organizations use generative AI?
It depends on the survey. McKinsey’s 2025 data showed 79% of respondents regularly using generative AI in at least one function, and its 2026 survey found nearly nine in ten using AI regularly. The US Census Bureau counted 19.8% of all US businesses using AI as of May 3, 2026, and Eurostat found 20.0% of EU enterprises with 10 or more employees in 2025.
How large is the generative AI market?
Estimates vary by scope. Gartner’s September 2026 forecast puts generative AI model spending at $28.3 billion in 2026, within $2.67 trillion of total AI spending. Menlo Ventures estimated $37 billion of US enterprise generative AI spending in 2025. Always check which categories a figure includes.
How fast is generative AI growing?
Gartner forecasts generative AI model spending to grow 117% in 2026, from $13.0 billion to $28.3 billion, and total AI spending to grow 49.5%. Gemini grew from 400 million monthly users in May 2025 to over 1 billion in August 2026, and ChatGPT weekly users rose from 400 million in February 2025 to 1.2 billion in September 2026.
How much are companies spending on generative AI?
Menlo Ventures estimated $37 billion of enterprise generative AI spending in 2025, up 3.2 times from 2024. In McKinsey’s 2026 survey, 28% of respondents spent more than 10% of their ICT budget on AI and 60% expected to increase AI investment. Gartner expects infrastructure to be the largest share of overall AI spending.
What is generative AI used for most?
Consumers mostly use it for practical guidance, information seeking and writing, which made up nearly 80% of ChatGPT conversations in an NBER study through mid-2025. At work, writing dominated ChatGPT use. Anthropic’s data shows coding is the largest category on Claude. Eurostat found EU firms most often use AI to analyze written language.
What percentage of employees use generative AI at work?
Gallup’s May 2026 survey of 22,573 US employees found 52% used AI at work at least a few times a year, 30% a few times a week or more, and 15% daily. Gallup’s question covers AI broadly, and only 47% said their organization had integrated AI tools.
Does generative AI improve productivity?
Named studies show gains in narrow tasks: 14% to 15% more issues resolved per hour in customer support, 26% more pull requests for Copilot users, and 50% more output in marketing. A METR trial found experienced developers 19% slower in 2025. In McKinsey’s 2026 survey, 80% said AI improved their own productivity.
What is the ROI of generative AI?
Enterprise-level returns are not yet widespread. McKinsey’s 2026 survey found 37% of respondents attribute any EBIT impact to AI, unchanged from 2025, and about 6% are high performers. Deloitte’s survey found 66% reporting productivity gains. Most ROI evidence is self-reported, so it is best treated as directional.
How many developers use AI coding tools?
Stack Overflow reported 79% of its 2025 respondents using AI tools in development, up from 62% in 2024, with 47.1% using them daily. A smaller April 2026 pulse survey found 59% using AI agents, up from 31% in 2025. Stack Overflow’s 2026 survey results were not yet published at the time of writing.
Will generative AI affect jobs?
Aggregate job losses have not appeared, but expectations are rising. McKinsey found 39% of respondents expect AI-related headcount declines next year, while only 14% reported one in the past year. Stanford reports employment for software developers aged 22 to 25 down nearly 20% from its 2022 peak. PwC found a 62% wage premium for AI skills.
What are the biggest barriers to generative AI adoption?
Cited barriers include unclear value for individual workers, operating costs (about 20% of McKinsey respondents say costs constrain use), weak governance, and the need to redesign workflows. IBM found 68% of breached organizations lacked AI governance policies, and shadow AI was involved in 43% of incidents.
Which countries have the highest generative AI adoption?
Microsoft’s telemetry-based measure showed the UAE at 70.1% of working-age people in Q1 2026, with Singapore next at over 60%. The Global North averaged 28.8% against 16.2% in the Global South in Q2 2026. In the EU, Denmark led enterprise AI use at 42.0% in 2025.
What do credible forecasts say about generative AI growth?
Gartner forecasts worldwide AI spending of $3.64 trillion in 2027 and generative AI model spending of $51.6 billion. IDC forecast about $1.3 trillion of AI spending by 2029 and $1.08 trillion of AI infrastructure spending in 2029. These are forecasts, and firms revise them often.
Key Takeaways
Consumer adoption is at internet scale, but platform counts (weekly versus monthly) and surveys (49% of US adults) are different measures.
Workplace use is broad but shallow: 52% of US employees use AI, while 30% use it frequently.
Organizational adoption depends on the yardstick: nearly nine in ten in McKinsey’s survey, about one in five in Census and Eurostat counts.
Scaling is rising faster than financial returns: 44% scaling AI, 37% reporting EBIT impact, both measured in 2026.
Large companies are pulling ahead, especially in agents (40% versus 22% scaling).
Money is concentrated in infrastructure and a few frontier labs, so total AI spend overstates generative AI software spend.
Productivity gains are best documented in structured tasks and are weaker for complex judgment work.
Governance is the clearest gap: shadow AI in 43% of studied incidents and 68% of breached organizations without AI governance.
Job effects so far show up in early-career hiring more than in aggregate unemployment.
Actionable Next Steps
Inventory current AI use, including unapproved tools, because IBM tied shadow AI to 43% of studied incidents.
Set a short acceptable-use policy and an approval path for tools and data types.
Pick two or three workflows with measurable outputs, such as support resolution, code review or marketing production, where named studies show gains.
Record a baseline for cycle time, quality and cost before launching, since McKinsey’s high performers more often define measurement processes.
Run a controlled pilot and track cost per task, including tokens, because 20% of McKinsey respondents report cost constraints.
Redesign the workflow around the tool; nearly three-quarters of high performers did, versus one-quarter of others.
Train staff by role and protect entry-level learning paths, given the pressure on early-career hiring.
Review build-versus-buy for software now that 32% of McKinsey respondents skipped a purchase to build in-house.
Scale only what shows measured results, and review forecasts and vendor claims against their publication dates.
Glossary
Generative AI — AI that creates text, images, code, audio or video from prompts.
Large language model (LLM) — A model trained on large text datasets to understand and generate language.
AI agent — Software that plans and takes multi-step actions toward a goal with limited human input.
AI copilot — An assistant embedded in a tool, such as an editor or office suite, that helps a user complete tasks.
Shadow AI — AI tools employees use without approval or oversight from IT or security.
AI governance — Policies, controls and accountability for how AI is selected, used and monitored.
Inference — Running a trained model to produce an answer; training is the process that builds the model.
EBIT — Earnings before interest and taxes, a measure of operating profit.
CAGR — Compound annual growth rate, the average yearly growth rate over a period.
Production deployment — Use of AI in live operations, as opposed to a pilot or proof of concept.
Telemetry — Usage data collected automatically from software, used by Microsoft to estimate AI diffusion.
Sources & References
Stanford HAI. “The 2026 AI Index Report, Chapter 4: Economy” Stanford University, April 2026.
Stanford HAI. “The 2026 AI Index Report” Stanford University, 2026.
McKinsey & Company. “The state of AI in 2026: On the road to ROI” McKinsey, August 25, 2026.
Gartner. “Gartner Forecasts Worldwide AI Spending to Grow 49.5% in 2026” Gartner, September 16, 2026.
Gartner. “Gartner Forecasts Worldwide AI Platforms and Models Market to Grow 63% in 2026” Gartner, July 20, 2026.
Gartner. “Gartner Forecasts Worldwide AI-Optimized IaaS Spending to Grow 96% Through 2026” Gartner, August 10, 2026.
Gartner. “Gartner Says Worldwide AI Spending Will Total $2.5 Trillion in 2026” Gartner, January 15, 2026.
Gartner. “Gartner Forecasts Worldwide AI Spending to Grow 47% in 2026” Gartner, May 19, 2026.
OpenAI. “DevDay 2026 Recap” OpenAI, September 2026.
SiliconANGLE. “Google’s Gemini AI app passes 1 billion monthly active users” SiliconANGLE, August 11, 2026.
Pew Research Center. “Americans and AI 2026: Chatbots, Smart Devices and Views on Impact (chart: chatbots used)” Pew Research Center, June 2026.
The Next Web. “Half of Americans now use AI chatbots (coverage of Pew)” The Next Web, June 17, 2026.
Microsoft AI Economy Institute. “The continued state of global AI diffusion in 2026” Microsoft, September 21, 2026.
Microsoft AI Economy Institute. “The state of global AI diffusion in 2026” Microsoft, May 7, 2026.
Redmondmag. “Microsoft Report: Global AI Use Rises as Adoption Gap Continues to Widen” Redmondmag, May 7, 2026.
Gallup. “Organizational AI Adoption Jumps Six Points” Gallup, Q2 2026 (survey May 6–20, 2026).
Gallup. “Rising AI Adoption Spurs Workforce Changes” Gallup, April 12, 2026 (survey Feb 4–19, 2026).
Gallup. “Global Indicator: Artificial Intelligence” Gallup, 2026.
The Register. “Workplace AI adoption falls flat” The Register, January 26, 2026.
U.S. Census Bureau. “Business Trends and Outlook Survey (BTOS)” Census Bureau, data through May 3, 2026.
U.S. Census Bureau. “The Microstructure of AI Diffusion (CES Working Paper 26-25)” Census Bureau, 2026.
Digital Watch. “US Census Bureau reports higher AI adoption among larger firms” Digital Watch, 2026.
Eurostat. “20% of EU enterprises use AI technologies” Eurostat, December 2025.
Menlo Ventures. “2025: The State of Generative AI in the Enterprise” Menlo Ventures, December 9, 2025.
Crunchbase News. “Global venture funding hits record $510 billion in H1 2026” Crunchbase, July 2, 2026.
AI Weekly. “US Venture Hits $412.7B in H1 2026, AI Takes 86% (PitchBook-NVCA data)” AI Weekly, 2026.
IDC. “AI Infrastructure Spending Holds Near $90 Billion in Q1 2026; 2026 Forecast Raised to $497 Billion” IDC, 2026.
IDC. “Agentic AI to Dominate IT Budget Expansion Over Next Five Years” IDC, August 26, 2025.
BizTech Reports. “Asia/Pacific AI and GenAI Spending to Reach $370 Billion by 2029 (IDC)” BizTech Reports, April 28, 2026.
Stack Overflow. “Getting ready for 2026 results: A look back on Developer Survey findings” Stack Overflow, September 30, 2026.
PwC. “2026 Global AI Jobs Barometer press release” PwC, June 15, 2026.
IBM. “IBM Study: One in Four Malicious Breaches are AI-Enabled (Cost of a Data Breach Report 2026)” IBM, July 29, 2026.
Deloitte AI Institute. “The State of AI in the Enterprise” Deloitte, 2026 edition.
Chatterji et al.. “How People Use ChatGPT (NBER Working Paper 34255)” NBER, September 2025.
Harvard D^3 Institute. “Here’s What People Actually Do With ChatGPT” Harvard, 2025.
Anthropic. “Anthropic Economic Index report: Learning curves” Anthropic, March 24, 2026.
