AI Usage Statistics 2026: Who Uses AI, How Often, For What

AI usage statistics for 2026 depend on who is asked and what is counted. Microsoft's telemetry puts the share of people aged 15 to 64 who have used a generative AI product at 18.8% worldwide (Microsoft AI Economy Institute, Sep 2026). In PwC's survey of 49,364 workers in 48 countries, 64% used AI at work in the past 12 months, up 10 points in a year (PwC, Sep 2026). The first counts everyone in an age band, the second only people in work, so the two numbers do not conflict.

The page covers consumers, workers, developers and companies: how many use AI, how often, which tools and tasks, which industries and company sizes lead, and how much use happens without the employer's approval. Each figure links to the organisation that collected it, and survey firms and vendors are named so you can weigh them. Wider market data sits on our AI statistics and generative AI statistics pages, and traffic from AI assistants is on AI search statistics.

Key Takeaways

25% of US adults use AI every day, up from 19% a year earlier (Menlo Ventures, Sep 2026).
Weekly use of generative AI across six countries nearly doubled in a year, from 18% to 34% of people (Reuters Institute, Oct 2025).
49% of US adults use AI chatbots and 51% do not (Pew Research Center, Oct 2026).
China passed 700 million generative AI users in the first half of 2026, more than half of its population (CNNIC, Oct 2026).
74% of frontline employees are now regular AI users, up 23 percentage points from 2025 (BCG, Jun 2026).
73% of developers use coding assistants and agents every day (Stack Overflow, Oct 2026).
AI use rises with company size: 17.6% of US businesses with 1 to 4 employees used AI in the prior two weeks, against 30.6% of those with 250 or more (U.S. Census Bureau, Feb 2026).
The median business on Ramp spends $11.38 per employee a month on AI, while the top 1% spend $7,449 (Ramp, Jun 2026).
47% of generative AI users at work still use personal AI apps, down from 78% a year earlier (Netskope, Jan 2026).
29% of employees have used unsanctioned AI agents for work tasks (Microsoft Security, Feb 2026).
One in five organisations reported a breach caused by shadow AI, and heavy shadow AI use added $670,000 to the average breach cost (IBM, Jul 2025).
Only 26% of AI users say their leadership is clearly and consistently aligned on AI (Microsoft Work Trend Index, May 2026).

How many people around the world use AI?

Global AI use is climbing about a point a quarter at the moment and is far from even. The spread runs from a majority of the population in the UAE to fewer than one person in ten in the lowest-ranked economies.

AI usage statistics: share of people aged 15 to 64 who have used generative AI, by economy, Q2 2026 — Figure 1: The map shades each economy by the share of people aged 15 to 64 who have used a generative AI product, and all ten of the top economies are high-income. Source: Microsoft AI Economy Institute, Global AI Diffusion Report, Sep 2026. Screenshot of the source's interactive page, cropped, captured 14 Oct 2026.

Use rose in every period Microsoft measured, from 15.1% of people aged 15 to 64 in the first half of 2025 to 17.8% in the first quarter of 2026 (Microsoft AI Economy Institute, Sep 2026).

The Global North reached 28.8% in the second quarter of 2026, against 16.2% in the Global South (Microsoft AI Economy Institute, Sep 2026).
The gap between the two grew from 9.8 percentage points in the first half of 2025 to 12.6 points (Microsoft AI Economy Institute, Sep 2026).
Menlo Ventures estimates that 2.0 billion people use AI worldwide, about 24% of the global population, up 11% in a year (Menlo Ventures, Sep 2026).
China had 602 million generative AI users at the end of 2025, up 141.7% in a year, a penetration of 42.8% (CNNIC, Mar 2026).
Across Argentina, Denmark, France, Japan, the UK and the US, the share of people who had ever used a standalone generative AI system rose from 40% in 2024 to 61% in 2025 (Reuters Institute, Oct 2025).
90% of people in those six countries had heard of at least one of 13 AI tools, up from 78% a year earlier (Reuters Institute, Oct 2025).

Microsoft's figure for China is 17.5%, far below CNNIC's count. The two use different methods, so compare each series only with itself.

Which countries use AI the most?

Wealthy, highly digital economies lead. The United Arab Emirates ranked first at 73.3% of people aged 15 to 64 in the second quarter of 2026, ahead of Singapore at 64.3% (Microsoft AI Economy Institute, Sep 2026).

Ireland (49.9%), France (49.6%) and Norway (49.4%) completed the top five (Microsoft AI Economy Institute, Sep 2026).
The United States ranked 21st at 33.0%, and Germany reached 32.3% (Microsoft AI Economy Institute, Sep 2026).
Japan ranked 45th at 24.7%, and Cambodia ranked last of 147 economies at 6.1% (Microsoft AI Economy Institute, Sep 2026).
South Korea made the largest quarterly gain, 3.5 points from 37.1% to 40.6%, and moved from 16th to 12th (Microsoft AI Economy Institute, Sep 2026).
Denmark scored 2.1 on Anthropic's usage index, meaning its residents use Claude at about twice the rate their share of the world's working-age population would predict (Anthropic, Jan 2026).
A 1% rise in GDP per capita goes with a 0.7% rise in Claude usage per capita across countries (Anthropic, Jan 2026).
The top 20 countries account for 48% of per-capita Claude usage, up from 45%, so the gap between countries is widening (Anthropic, Mar 2026).

Who uses AI, and who holds out?

Age separates users from non-users more than anything else. 66% of US adults aged 18 to 29 use AI chatbots, against 23% of those aged 65 and older (Pew Research Center, Oct 2026).

61% of adults aged 30 to 49 and 42% of those aged 50 to 64 use chatbots (Pew Research Center, Oct 2026).
70% of Asian adults use chatbots, against 46% of White adults (Pew Research Center, Oct 2026).
77% of consumers in households earning $100,000 or more use AI, against 56% in households earning under $50,000 (Menlo Ventures, Sep 2026).
84% of US parents use AI, against 56% of non-parents (Menlo Ventures, Sep 2026).
Among US adults who never use chatbots, 83% say lack of interest is a reason and 79% cite how their personal information would be used (Pew Research Center, Oct 2026).
67% of non-users say they are unlikely to use a chatbot in the next 12 months, and 5% say they are very or extremely likely to (Pew Research Center, Oct 2026).
70% of US non-users of AI distrust AI-generated information, up from 58% a year earlier (Menlo Ventures, Sep 2026).

How often do people use AI, and with which tools?

Frequency turns a trial into a habit, and daily AI use now passes ride-sharing and food delivery combined.

Share of US adults who use consumer AI tools, ride-sharing, food delivery and streaming daily, 2026 — Figure 2: Daily use of consumer AI tools is already above ride-sharing and food delivery combined, and only streaming is more common among the habits Menlo tracks. Source: Menlo Ventures, The State of Consumer AI 2026, Sep 2026. Chart published by the source, captured 14 Oct 2026.

How many people use AI every day?

Among US adults who use AI, 39% use it daily, up from 31% a year earlier (Menlo Ventures, Sep 2026).

General AI assistants alone are used daily by 22% of consumers, more than ride-sharing (6%) and food delivery (7%) combined (Menlo Ventures, Sep 2026).
Of the behaviours Menlo tracks, only streaming, at 46%, is substantially more widespread than daily AI use (Menlo Ventures, Sep 2026).
24% of AI users spend more than an hour a day with AI and 10% spend more than two, while 29% spend under 15 minutes (Menlo Ventures, Sep 2026).
52% of AI users say they use AI more than they did a year ago (Menlo Ventures, Sep 2026).
Millennials lead daily use at 37%, and daily use among parents rose from 29% to 40% (Menlo Ventures, Sep 2026).
Of the developers who use AI daily, 31% use it for more than four hours a day (Stack Overflow, Oct 2026).

How many AI tools does the average user have?

Three. The average AI user now uses 3.0 general AI assistants, up from 2.2 a year earlier (Menlo Ventures, Sep 2026).

96% of AI users have used a general AI assistant, up from 92% (Menlo Ventures, Sep 2026).
Gemini's reach among AI users grew from 36% to 58% in a year, and Claude's from 7% to 20% (Menlo Ventures, Sep 2026).
72% of AI users use both general and specialised AI products, up from 60% (Menlo Ventures, Sep 2026).
55% of AI users pay for at least one AI product, and the 14% who spend $100 or more a month account for 60% of consumer AI spending (Menlo Ventures, Sep 2026).
Across organisations, Gemini was used in 69% and Microsoft 365 Copilot in 52%, while Perplexity rose from 23% to 35% (Netskope, Jan 2026).
43.8% of US businesses on Ramp paid Anthropic in August 2026, against 39.8% for OpenAI (Ramp, Sep 2026).
A record 8% of firms switched AI model providers in September 2026, on a three-month average (Ramp, Oct 2026).

How do people talk to AI and let it act for them?

More people now speak to AI than type to it some of the time: 46% of AI users talk to AI at least some of the time, and for 19% voice is the main way (Menlo Ventures, Sep 2026).

Alexa usage fell from 32% to 22% and Siri usage from 25% to 20% (Menlo Ventures, Sep 2026).
Weekly use splits between phone apps (56%) and computers (54%), and wearables reach 15% of AI users weekly (Menlo Ventures, Sep 2026).
41% of AI users have tried an AI agent and 24% use one regularly (Menlo Ventures, Sep 2026).
32% of AI users have let AI act on their behalf without final approval (Menlo Ventures, Sep 2026).
Agent users have given agents access to their email (36%), web browsers (33%) and messaging apps (31%) (Menlo Ventures, Sep 2026).
Access to health apps (23%) and financial accounts (20%) is granted far less often (Menlo Ventures, Sep 2026).
Codex leads among agent users at 35%, ahead of Claude Code and Cowork at 32% and Perplexity at 24% (Menlo Ventures, Sep 2026).

What do people use AI for?

The most common use of AI is still asking it a question. Making things comes second, and both are moving into areas where people used to rely on a professional, such as health advice and buying decisions.

Share of US adults who use AI for each activity, among people who do the activity, 2025 and 2026 — Figure 3: AI is now the default for six activities, up from one a year earlier, with image creation rising from 34% to 53% in a single year. Source: Menlo Ventures, The State of Consumer AI 2026, Sep 2026. Chart published by the source, captured 14 Oct 2026.

What do consumers ask AI to do?

Finding information is the top weekly use across six countries, at 24% of people, more than double the 11% of a year earlier (Reuters Institute, Oct 2025).

Creating media is second at 21% weekly, up 7 points (Reuters Institute, Oct 2025).
26% of US adults used AI to research a topic of personal interest, the most common AI activity Menlo measured (Menlo Ventures, Sep 2026).
Among people who do each activity, 61% use AI for writing support and 58% for coding projects (Menlo Ventures, Sep 2026).
Using AI to answer health questions rose from 14% to 25% of US adults, and for mental health support from 8% to 14% (Menlo Ventures, Sep 2026).
Of shoppers who use AI, 50% compared prices with it and 43% used it to narrow a shortlist (Menlo Ventures, Sep 2026).
37% of those shoppers made a final purchase decision on its recommendation, and 26% let AI complete a purchase (Menlo Ventures, Sep 2026).
In China, 76.0% of generative AI users ask it questions, 47.8% process images or video and 37.6% process text (CNNIC, Oct 2026).

Shopping is now an AI task for a large share of AI users, so any business that sells online has a reason to know whether assistants recommend it.

What work tasks do people hand to AI?

Writing and research dominate. Among US employees who use AI at work, 51% use it for writing and editing and 49% for search or research (Gallup, Jul 2026).

39% use AI for general assistance and problem solving (Gallup, Jul 2026).
Frequent users are far more likely than infrequent users to code with AI (22% against 8%) and to automate work (21% against 8%) (Gallup, Jul 2026).
Among US businesses whose employees use generative AI, 85.4% use it for writing or editing and 49.9% for searching for information or technical help (U.S. Census Bureau, Feb 2026).
Only 12.7% of those businesses use it for software coding or debugging, and 15.5% for customer support (U.S. Census Bureau, Feb 2026).
In a week of Microsoft 365 Copilot chats, 49% supported cognitive work such as analysis and problem solving, and 15% were about finding information (Microsoft Work Trend Index, May 2026).
Claude work conversations most often produce documents and reports (20%), explanations (9%) and email drafts (7%) (Anthropic, Jun 2026).
52% of Claude.ai conversations were augmentation, where the person works with the model, and 45% were automation, where the model completes the task (Anthropic, Jan 2026).

How do developers use AI?

Coding assistants and agents are the dominant AI use case among developers at 66%, ahead of general-purpose chatbots at 63% (Stack Overflow, Oct 2026).

69.3% of developers use AI to generate code in a familiar area and 63.8% to debug, troubleshoot or refactor (Stack Overflow, Oct 2026).
85% of developers regularly use AI tools for coding, and 15% have not adopted them (JetBrains, Oct 2025).
68% of developers expect their employers to require proficiency in AI tools (JetBrains, Oct 2025).
More than 1.1 million public GitHub repositories import an LLM SDK, up 178% in a year (GitHub, Oct 2025).
Nearly 80% of new GitHub developers use Copilot in their first week (GitHub, Oct 2025).
Programming grew from about 11% of token volume on OpenRouter in early 2025 to more than 50% in recent weeks (OpenRouter, Dec 2025).
Computer and mathematical tasks account for 35% of Claude.ai conversations (Anthropic, Mar 2026).

How do workers use AI on the job?

Workers use AI, and the heaviest users save a workday a week. The gains are uneven by seniority and function, and most of the time saved is not pointed at anything.

Share of respondents who save at least eight hours a week with AI, by role and by job function, 2026 — Figure 4: Leaders save the most time with AI, 60% of them at least a workday a week against 42% of frontline employees, and marketing and IT lead the functions. Source: BCG, AI at Work: Strategy Matters More Than Tools, Jun 2026. Chart published by the source, captured 14 Oct 2026.

22% of workers now use generative AI daily, up from 14% a year earlier (PwC, Sep 2026).

59% of workers expect their use of AI tools to increase over the next 12 months (PwC, Sep 2026).
51% of PwC's AI-enabled front-runner workers use generative AI daily, against 11% of the core workforce (PwC, Sep 2026).
The majority of the core workforce say they have never used AI (PwC, Sep 2026).
58% of employees in KPMG's 47-country study intentionally use AI at work, and 31% use it weekly or daily (KPMG and University of Melbourne, Apr 2025).
Where AI is available, 67% of leaders use it a few times a week or more, against 46% of individual contributors (Gallup, Apr 2026).
Frequent use is 88% among employees who strongly agree AI integrates well with their work systems, against 55% of others (Gallup, Apr 2026).

Which jobs and industries use AI the most?

Finance and the public sector lead, and retail trails. 40% of finance employees use AI frequently, against 19% in retail (Gallup, Mar 2026).

43% of public sector employees use AI at least a few times a year, and 21% use it frequently (Gallup, Mar 2026).
Private sector use is 41% at least a few times a year and 25% frequently (Gallup, Mar 2026).
49% of jobs have seen at least a quarter of their tasks performed using Claude (Anthropic, Mar 2026).
In Microsoft's telemetry, software and technology (16%), manufacturing (13%), financial institutions (11%) and retail (9%) lead the industries using AI agents (Microsoft Security, Feb 2026).
India and the Middle East lead frontline AI adoption, while the US, France and Italy trail (BCG, Jun 2026).

How much time does AI save workers?

Leaders gain the most. 60% of leaders who are regular AI users save at least a workday a week, against 42% of frontline employees (BCG, Jun 2026).

Frontline employees save a workday a week most often in marketing (60%), IT (53%) and human resources (50%) (BCG, Jun 2026).
66% of frontline employees get limited or no guidance on what to do with the time saved, and 58% do not reinvest it in more strategic work (BCG, Jun 2026).
66% of AI users say AI lets them spend more time on high-value work, and 58% say they produce work they could not have a year ago (Microsoft Work Trend Index, May 2026).
Claude users report productivity gains in speed (86%), scope (82%) and quality (69%) (Anthropic, Jun 2026).
45% of employees who use AI for one or two purposes say it has had a positive impact on their productivity, rising to 66% for three or four purposes and 90% for seven or more (Gallup, Jul 2026).

How many companies use AI, and how deeply?

Company AI use looks small in government surveys and large in spending data. Census asks whether a business used AI in the last two weeks, while Ramp counts who pays for it, and the gap between the two says as much about measurement as about adoption.

Share of US businesses with spending on AI models, subscriptions, tokens and APIs, with the Census estimate, 2023 to 2026 — Figure 5: Ramp's spending-based index runs at 56.1% of businesses while the Census survey estimate sits at 22.1%, a gap of more than two to one. Source: Ramp, Ramp AI Index, Oct 2026. Screenshot of the source page, cropped, captured 14 Oct 2026.

Which companies use AI the most?

Bigger companies use AI more. In the Census Bureau's AI supplement, 18.5% of businesses with 20 to 49 employees used AI in the prior two weeks, against 25.3% of those with 100 to 249 (U.S. Census Bureau, Feb 2026).

In the prior six months, employees used generative AI at 19.1% of businesses with 1 to 4 employees and at 38.7% of those with 250 or more (U.S. Census Bureau, Feb 2026).
74% of tech businesses and 60% of finance businesses on Ramp paid for AI in October 2025 (Ramp, Nov 2025).
Retail businesses on Ramp reached 36%, having grown fastest in the month (Ramp, Nov 2025).
Almost 60% of small businesses use AI for operations, more than double the share in 2023 (U.S. Chamber of Commerce, Aug 2025).
More than 80% of Fortune 500 companies use active AI agents built with low-code or no-code tools (Microsoft Security, Feb 2026).
Among businesses with no plan to use AI, 61.6% say it is not applicable to them, 22.0% cite lack of knowledge of what AI can do and 20.7% cite privacy or security (U.S. Census Bureau, Feb 2026).

What do companies use AI for?

Sales and marketing is the most common business function for AI, used by 14.3% of US businesses in the six months to early 2026 (U.S. Census Bureau, Feb 2026).

Strategy and business development (12.4%), IT (11.4%) and research and development (11.2%) follow (U.S. Census Bureau, Feb 2026).
Customer service (7.6%), finance and accounting (6.6%) and human resources (4.2%) trail, and distribution is lowest at 0.9% (U.S. Census Bureau, Feb 2026).
Among AI-using businesses, 43.7% used AI to supplement a task done by an employee, 10.1% to perform a task an employee used to do, and 10.6% to introduce a new task (U.S. Census Bureau, Feb 2026).
70.9% of businesses that handed tasks to AI say they handed over only a small number (U.S. Census Bureau, Feb 2026).
95.7% of AI-using businesses report no change in total employment from AI, 2.3% an increase and 2.0% a decrease (U.S. Census Bureau, Feb 2026).
Among businesses planning to use AI, 62.8% expect to use it in sales and marketing (U.S. Census Bureau, Feb 2026).

How deeply do companies use AI, and what does it cost?

Most companies are still at the first rung. 78% of respondents say their organisation supports adoption of generative AI tools, 57% are redesigning workflows end to end, and 42% are building new business models with AI, up from 22% (BCG, Jun 2026).

64.3% of AI-using US businesses made no changes at all to use AI, and 15.4% developed new workflows (U.S. Census Bureau, Feb 2026).
The top 10% of businesses on Ramp spend $611 per employee a month on AI (Ramp, Jun 2026).
Spending per employee at the top 1% fell 9.7% in August 2026, to $7,205 (Ramp, Sep 2026).
The effective price per million tokens fell 41% to $0.68 (Ramp, Sep 2026).
Only 6.4% of AI-spending businesses on Ramp use open-source models (Ramp, Sep 2026).
70% of organisations connect to api.openai.com, 54% to AssemblyAI and 30% to Anthropic's APIs (Netskope, Jan 2026).

How much AI use happens without approval?

Employees adopted AI before their employers organised it, and the evidence sits in security logs. The picture is improving, but demand is growing faster than control.

Share of generative AI users by account type, personal, organisation and both, October 2024 to October 2025 — Figure 6: The share of generative AI users on personal accounts fell from about three-quarters to under half in a year as companies rolled out managed accounts, but a growing group still switches between the two. Source: Netskope, Cloud and Threat Report 2026, Jan 2026. Chart published by the source, covering October 2024 to October 2025, captured 14 Oct 2026.

How many employees use AI their employer has not approved?

78% of AI users bring their own AI tools to work, and the figure is 80% at small and medium-sized companies (Microsoft and LinkedIn, May 2024).

52% of people who use AI at work are reluctant to admit using it for their most important tasks (Microsoft and LinkedIn, May 2024).
Organisation-managed accounts rose from 25% to 62% of generative AI users in a year (Netskope, Jan 2026).
9% of users switch between personal and enterprise accounts, up from 4% (Netskope, Jan 2026).
The number of people using SaaS generative AI apps tripled in the average organisation (Netskope, Jan 2026).
Prompts sent to those apps rose sixfold, from 3,000 to 18,000 a month (Netskope, Jan 2026).
Netskope tracks more than 1,600 generative AI apps, up from 317, while the average organisation uses 8, up from 6 (Netskope, Jan 2026).

What company data goes into AI tools?

In the average organisation, 3% of generative AI users commit data policy violations, averaging 223 incidents a month, and incident counts doubled in a year (Netskope, Jan 2026).

Source code accounts for 42% of generative AI data policy violations, regulated data for 32% and intellectual property for 16% (Netskope, Jan 2026).
In the top quartile of organisations the average is 2,100 incidents a month, spread across 13% of generative AI users (Netskope, Jan 2026).
60% of insider threat incidents involve personal cloud app instances (Netskope, Jan 2026).
90% of organisations actively block at least one generative AI app, and ZeroGPT (45%) and DeepSeek (43%) are blocked most often (Netskope, Jan 2026).
Shadow AI incidents compromised personal data in 65% of cases and intellectual property in 40%, against 53% and 33% across all breaches (IBM, Jul 2025).
13% of organisations reported a breach of an AI model or application, and 97% of them lacked AI access controls (IBM, Jul 2025).

How many companies have AI training and policies?

Few. Only 47% of employees in KPMG's global study have received AI training, and 40% say their workplace has a policy or guidance on generative AI use (KPMG and University of Melbourne, Apr 2025).

88% of BCG's respondents expect to need major upskilling in the next five years, and 36% feel properly trained (BCG, Jun 2026).
Only a third of frontline employees say leaders communicate clearly about AI (BCG, Jun 2026).
50% of organisations lack enforceable data protection policies for generative AI apps (Netskope, Jan 2026).
63% of breached organisations had no AI governance policy or were still developing one, and only 37% had policies to manage AI or detect shadow AI (IBM, Jul 2025).
Among organisations with a governance policy, only 34% audit regularly for unsanctioned AI (IBM, Jul 2025).
15.0% of AI-using US businesses trained current staff to use AI, and 1.3% hired staff trained in AI (U.S. Census Bureau, Feb 2026).
Workers rate their access to learning and development resources at 51%, down from 59% a year earlier (PwC, Sep 2026).
50% of chief human resources officers are not confident in managers' ability to guide AI use (Gallup, Aug 2026).

Methodology and Sources

We included a number only after tracing it to the organisation that produced it and opening that page to confirm the figure and its date. Figures quoted in roundups, including competitor statistics pages, were followed back to the original study or dropped. Most numbers come from statistical agencies, company research teams and survey firms (Gallup, Pew, PwC, BCG, KPMG, Microsoft, Menlo Ventures), which are named in each citation so you can judge the sample. Several sources measure different groups: US adults, workers, developers, businesses on a payments platform, or customers of one AI product. We never blend them into one claim, and where a source covers a single assistant, as Anthropic's Economic Index does, the text says so. Figures from the Census Bureau's AI supplement were calculated from the published tables for the period 17 November 2025 to 8 February 2026. We refresh the page monthly.

Sources

Glossary

Generative AI: AI that produces text, images, code or audio in response to a prompt.
LLM: large language model, the type of model behind most generative AI tools.
AI agent: an AI system that carries out multi-step tasks on its own, such as running code or filling in forms.
Shadow AI: AI tools used for work without the employer's approval or knowledge.
BYOAI: bring your own AI, meaning employees use personal AI tools or accounts for work.
Global North and Global South: Microsoft's grouping of economies by income and development level in its diffusion index.
Telemetry: usage data collected automatically from software, in Microsoft's case aggregated and anonymised.
BTOS: the Census Bureau's Business Trends and Outlook Survey of US employer businesses.
Token: a small unit of text that language models read and write; usage and prices are measured per million tokens.
API: application programming interface, the way software connects to an AI model without a chat window.
DLP: data loss prevention, tools that detect and block sensitive data leaving an organisation.
CHRO: chief human resources officer.

147 AI Usage Statistics October 2026