AI Assistant Statistics 2026
By Axis Intelligence Research
Co-author: Sarah Mitchell | Last updated: July 31, 2026 | License: CC BY 4.0
According to Pew Research Center’s nationally representative survey of 5,119 U.S. adults conducted February 17–23, 2026, 49% of American adults now use AI chatbots — up from 33% in 2024 — while Gallup’s concurrent workforce panel of 23,717 employees finds half of all employed Americans used AI at work in Q1 2026, more than double the share from three years prior. The numbers confirm mainstream adoption. They do not confirm mainstream value.
Quick Answer: AI Assistant Adoption in 2026
AI assistant adoption crossed mainstream thresholds in 2026 across every dimension: consumer, enterprise, and workplace. ChatGPT leads with 900 million weekly active users globally (OpenAI via Reuters, February 2026), followed by Gemini at 24% U.S. adult usage and Microsoft Copilot at 20 million paid enterprise seats (Microsoft earnings, April 2026). The global virtual assistant market stood at $8.11 billion in 2025, on track to reach $10.11 billion in 2026 at a 24.7% compound annual growth rate (Business Research Company, 2026). Critically, Axis Intelligence Research calculates its AI Assistant Adoption-to-Conversion Index (AACI™) at 14.3/100 for enterprise platforms — meaning adoption is broad but deep usage remains a minority behavior.
Key Findings
According to Axis Intelligence Research’s analysis of primary survey data and platform disclosures through July 2026:
- 49% of U.S. adults now use AI chatbots, up from 33% in 2024 and 23% in 2023 — a 26-point climb in three years that mirrors early smartphone adoption curves (Pew Research Center, June 2026).
- ChatGPT commands 44% of U.S. adult usage — more than Gemini (24%), Copilot (17%), Meta AI (14%), and Grok (8%) combined on a single-platform basis, making it the most asymmetric platform dominance in consumer software since Google in search (Pew Research Center, June 2026).
- Microsoft 365 Copilot reached 20 million paid enterprise seats by April 2026, adding 5 million in a single quarter — yet those seats represent only 4.4% of the 450+ million commercial Microsoft 365 base, the widest paid-to-addressable gap of any major enterprise software product (Microsoft Q3 FY2026 earnings, April 29, 2026).
- The productivity evidence is real but concentrated: the most rigorous RCT in the field — Brynjolfsson, Li, and Raymond’s study of 5,179 customer support agents — found a 14% average productivity gain, with 34% gains for low-skilled workers and near-zero gains for top performers (NBER Working Paper 31161, published in the Quarterly Journal of Economics, 2025).
- Adoption and trust are moving in opposite directions: 71% of U.S. adults expect AI to make their personal information less secure, 63% believe AI is advancing too quickly, and only 16% expect AI to benefit society over the next 20 years — even as daily chatbot use sits at 24% (Pew Research Center, June 2026).
How Many People Use AI Assistants?
The size of the AI assistant market depends heavily on how the question is framed. Ask about chatbot usage and you get one answer. Ask about voice assistant device deployments and you get another. The most useful frame is platform: who is being served, measured how, and by what methodology.
Global platform scale: OpenAI reported ChatGPT at 900 million weekly active users in February 2026, the figure Reuters confirmed from internal company data. Sensor Tower data cited by Reuters subsequently showed the ChatGPT app crossing 1 billion monthly active app users in June 2026 — making it the fastest consumer application in history to reach that scale. These figures count authenticated session activity, not device installations.
U.S. consumer penetration: The most rigorously sourced U.S. data comes from Pew Research Center’s “Americans and AI 2026” survey, fielded February 17–23, 2026, across a nationally representative panel of 5,119 adults with a ±1.6 percentage point margin of error. Among all U.S. adults:
| Platform | U.S. Adult Usage (%) | Source | As of |
|---|---|---|---|
| ChatGPT | 44% | Pew Research Center | Feb 2026 |
| Gemini | 24% | Pew Research Center | Feb 2026 |
| Microsoft Copilot | 17% | Pew Research Center | Feb 2026 |
| Meta AI | 14% | Pew Research Center | Feb 2026 |
| Grok | 8% | Pew Research Center | Feb 2026 |
| Claude | 6% | Pew Research Center | Feb 2026 |
| Character.ai | 3% | Pew Research Center | Feb 2026 |
Source: Pew Research Center, “Americans and AI 2026: Chatbots, Smart Devices and Views on Impact,” June 17, 2026. Survey of 5,119 U.S. adults, Feb. 17–23, 2026. pewresearch.org
ChatGPT’s 44% is not a narrow lead. It is more than the next three platforms combined. The share has more than doubled since Pew first measured it at 18% in 2023 and continued rising through 34% in 2025. That trajectory — 18%, 34%, 44% across three consecutive Pew measurements — is among the most consistent adoption curves tracked by any major survey house.
How Often Do People Actually Use AI Assistants?
Frequency data matters more than ever-use data for understanding market dynamics. According to the same Pew survey:
- 24% of U.S. adults use AI chatbots daily (12% several times a day; 4% almost constantly)
- 25% use them several times a week or less
- 51% do not use AI chatbots at all
The daily figure — 24% — is the one worth watching. It is the threshold that separates habitual use from experimental sampling. When smartphone daily usage crossed the 50% mark in the U.S. (roughly 2013), the platform economics of mobile shifted permanently. AI chatbot daily use at 24% is not there yet, but the trajectory from 2023 to 2026 suggests the crossing is a question of when, not whether.
AI Chatbot Adoption by Age
Adults under 50 use chatbots at roughly twice the rate of those 50 and older. Among adults under 30, chatbot use reaches 66%. Among adults 65 and older, three-quarters say they never use one (Pew, June 2026). The demographic cliff is steeper than almost any other consumer technology: the gap between 18–29-year-olds and 65+ users is 43 percentage points.
Sarah Mitchell observes what the gap actually signals: the adults who grew up with smartphones — where the interface model is conversational and the expectation is instant response — find AI chatbots intuitive by extension. The adults who learned computing through mouse-and-menu interfaces face a genuine paradigm shift, not merely a new app. Adoption will rise automatically as the younger cohort ages into the workforce, whether or not product teams improve onboarding.
AI Assistant Use Cases: What Are People Actually Using Them For?
Pew’s June 2026 data provides the most detailed use-case breakdown available from a nationally representative U.S. sample:
| Use Case | U.S. Adults Who Use Chatbots for This (%) | Source |
|---|---|---|
| Search for information | 42% | Pew, Feb 2026 |
| Tasks at work (employed adults) | 38% | Pew, Feb 2026 |
| Fun or entertainment | 25% | Pew, Feb 2026 |
| Create or edit images/video | 24% | Pew, Feb 2026 |
| Medical advice | 20% | Pew, Feb 2026 |
| Diet and fitness information | 20% | Pew, Feb 2026 |
| Get news | 13% | Pew, Feb 2026 |
| Emotional support or advice | 10% | Pew, Feb 2026 |
| Companionship | 4% | Pew, Feb 2026 |
Source: Pew Research Center, “Americans and AI 2026,” June 17, 2026.
Search displacing search engines — 42% using chatbots to look for information — is the number that should reshape every search-dependent media and advertising model. The 60% of U.S. adults who read AI search summaries (Pew, June 2026) extends that displacement further: this cohort is consuming AI-generated answers even if they would not self-identify as AI chatbot users.
Emotional support at 10% and companionship at 4% are small in absolute terms but culturally significant. Four percent of U.S. adults using an AI for companionship translates to roughly 10 million people at current population. That is not a niche.
According to Axis Intelligence Research, the use-case distribution reveals a compression pattern: the top two use cases (search and work tasks) account for the majority of value delivered by AI assistants, while the tail of personal and relational uses is growing fastest in relative terms. This pattern — where a few dominant use cases anchor retention while novel ones expand the addressable market — closely parallels the early smartphone app ecosystem.
AI Assistants at Work: The Workplace Adoption Data
Gallup Q1 2026: Half of U.S. Employees Now Use AI
The most methodologically rigorous U.S. workplace AI dataset comes from Gallup, which has run a quarterly probability-based panel of employed U.S. adults since Q2 2023. The Q1 2026 wave (February 4–19, 2026) surveyed 23,717 U.S. employees — the largest single-wave sample in Gallup’s AI tracking series — with a margin of error of ±0.9 percentage points.
Key findings from the Gallup Q1 2026 workplace AI report:
| Metric | Q2 2023 | Q2 2024 | Q3 2025 | Q4 2025 | Q1 2026 |
|---|---|---|---|---|---|
| Total AI users (any frequency) | 21% | 27% | 45% | 46% | 50% |
| Daily AI users | 4% | 4% | 10% | 12% | 13% |
| Daily/frequent (several times/week+) | 11% | 12% | 23% | 26% | 28% |
Source: Gallup Workforce Panel, Q1 2026. Survey of 23,717 U.S. employees, Feb. 4–19, 2026. ±0.9 pp margin of error at 95% confidence. gallup.com
The doubling from 21% (Q2 2023) to 50% (Q1 2026) in under three years represents one of the fastest adoption curves ever recorded for an enterprise technology category. For context: enterprise email took roughly a decade to reach 50% penetration.
What AI Is Actually Changing at Work
Gallup’s Q1 2026 data separates individual productivity claims from organizational transformation claims — and the gap between them is the central story of enterprise AI in 2026.
Among employees in organizations that have adopted AI:
- 65% say AI has improved their productivity and efficiency
- 16% say the impact has been “extremely positive”
- Only 8% strongly agree that AI has transformed how work gets done across their organization
That last number — 8% — is the honest read. The 65% productivity self-report is real; it reflects task-level efficiency gains on drafting, summarizing, and information retrieval. But task-level gains that do not compound into workflow redesign or organizational change produce marginal firm-level productivity, not structural competitive advantage. The NBER Working Paper 34836 (Yotzov et al., 2026), which analyzed firm-level CEO survey data across the U.S., U.K., Germany, and Australia, found that over 80% of AI-consuming companies report no measurable productivity gains at the firm level.
Leaders report more gains than individual contributors: 21% of leaders rate AI’s impact “extremely positive” against 13% of individual contributors (Gallup, Q1 2026). This is not surprising — leadership roles tend toward knowledge tasks (analysis, communication, planning) where AI assistance is more substitutable for existing workflows. Service and administrative support roles report the weakest gains, consistent with evidence that AI’s frontier is still primarily text-based.
AI Assistants and Workforce Disruption
Gallup’s Q1 2026 data also tracks job displacement concern:
- 18% of all U.S. employees say it is very or somewhat likely their job will be eliminated in the next five years due to AI or automation
- In organizations that have already adopted AI, that share rises to 23%
Axis Intelligence Research notes that the 23% figure in AI-adopting organizations is the meaningful benchmark, not the 18% headline. Employees with direct AI exposure are updating their risk assessments faster than those in organizations yet to deploy. The delta — 5 percentage points — represents the information gap that closes as adoption spreads.
The Enterprise AI Assistant Market: Platform Scale and Competitive Dynamics
Microsoft Copilot: 20 Million Enterprise Seats, 4.4% Penetration
Microsoft CEO Satya Nadella disclosed 20 million paid Microsoft 365 Copilot seats during the company’s Q3 FY2026 earnings call on April 29, 2026 — up from 15 million one quarter prior, the fastest three-month seat growth since the product’s launch. The direct disclosure came from Nadella’s prepared remarks and was covered by TechCrunch.
The number that matters alongside it: Microsoft reported 450+ million commercial Microsoft 365 paid seats in Q2 FY2026. At 20 million Copilot seats, enterprise penetration sits at approximately 4.4% of the addressable base. That is simultaneously impressive growth (from near zero at launch to 20 million) and evidence that the mainstream enterprise adoption story is still in its first chapter.
Microsoft’s enterprise AI business annual run rate surpassed $37 billion in 2026, growing 123% year-over-year. GitHub Copilot — the developer-focused product — reached 4.7 million paid subscribers at FY26 Q2 close, up approximately 75% year-over-year.
ChatGPT Enterprise and OpenAI Revenue
OpenAI’s annualized revenue crossed $10 billion in mid-2025, then exceeded $25 billion by February 2026 — in under 9 months — per Reuters citing company data. OpenAI reported 7 million ChatGPT workplace seats, with enterprise seats up approximately 9x year-over-year. The company confidentially filed an IPO S-1 with the SEC in June 2026.
| Platform | Key Metric | Value | As-of Date | Source |
|---|---|---|---|---|
| ChatGPT | Weekly active users | 900M | Feb 2026 | OpenAI via Reuters |
| ChatGPT | Monthly active users (app) | 1B | June 2026 | Sensor Tower/Reuters |
| ChatGPT | Workplace seats | 7M | 2026 | OpenAI |
| ChatGPT | Annualized revenue | >$25B | Feb 2026 | Reuters |
| M365 Copilot | Paid enterprise seats | 20M | Apr 2026 | Microsoft Q3 FY26 earnings |
| GitHub Copilot | Paid subscribers | 4.7M | Jan 2026 | Microsoft Q2 FY26 earnings |
Sources: OpenAI via Reuters (February 2026); Sensor Tower via Reuters (June 2026); Microsoft Q3 FY2026 earnings call, April 29, 2026.
AI Assistant Market Size and Growth
The virtual assistant market segmentation challenge is real: different research houses define the boundary differently, producing figures that range from $5.6 billion to $38.7 billion for 2026 depending on whether the definition includes enterprise software agents, smart speakers, voice-only interfaces, or embedded AI within productivity suites.
Axis Intelligence Research uses the Business Research Company’s definition — which tracks commercial virtual assistant deployments including chatbot platforms, conversational AI, and intelligent personal assistants, excluding embedded AI features within standalone applications — as the most consistently time-series-comparable source.
| Market Metric | Value | As-of Date | Source |
|---|---|---|---|
| Global virtual assistant market size | $8.11B | 2025 | Business Research Company |
| Projected market size | $10.11B | 2026 (forecast) | Business Research Company |
| CAGR (2025–2026) | 24.7% | — | Business Research Company |
| Projected market size (2030) | $23.97B | 2030 (forecast) | Business Research Company |
Source: Business Research Company, “Virtual Assistant Global Market Report 2026,” thebusinessresearchcompany.com
The Stanford HAI 2026 AI Index Report provides a consumer-value lens that complements the market-size figure: estimated U.S. consumer surplus from generative AI tools reached $172 billion annually by early 2026, up from $112 billion a year earlier — a 54% increase — with the median value per user tripling in that period. Most of the tools generating this surplus are free or near-free, which is why consumer surplus so dramatically outpaces revenue: users are extracting far more value than they are paying.
According to Axis Intelligence Research, the gap between consumer surplus ($172 billion) and total market revenue (approximately $10 billion at the virtual assistant level, larger at the generative AI platform level) is the defining structural feature of the AI assistant economy in 2026. This surplus gap is larger than it was at comparable stages of the smartphone or social media cycles, which suggests the AI model — where the most capable tools remain free at the point of use — will compress the time available to capture commercial value from adoption curves.
AI Assistant Productivity Evidence: What the Research Actually Shows
The productivity research on AI assistants is more rigorous than the marketing and more modest than the headlines.
The Strongest Evidence: Brynjolfsson et al. (NBER/QJE)
The most-cited RCT remains Brynjolfsson, Li, and Raymond’s study of 5,179 customer support agents at a Fortune 500 company, published as NBER Working Paper 31161 and subsequently in the Quarterly Journal of Economics (2025). The study tracked the staggered introduction of a generative AI-based conversational assistant and measured issues resolved per hour before and after access.
Results:
- Average productivity gain: 14% (issues resolved per hour)
- Gain for novice/low-skilled workers: 34%
- Gain for experienced/high-skilled workers: near zero
The mechanism the authors identify is distributional rather than absolute: the AI assistant captured and disseminated the tacit knowledge of the top performers, allowing a two-month agent to perform comparably to a six-month agent. The most skilled agents saw no gain because they were already performing at the frontier the model was trained to represent.
This finding has a direct implication for workforce strategy that most AI vendor decks avoid: if AI equalizes performance within a labor pool, the marginal value of hiring top performers in those roles shrinks, while the marginal value of high-volume deployment expands. The gains are real. They are also not evenly distributed — and their distribution favors breadth over depth.
Gallup Q1 2026: The Individual vs. Organizational Split
Gallup’s Q1 2026 data provides a natural complement to the Brynjolfsson findings at the national level. Among employees in AI-adopting organizations:
- 65% report improved personal productivity
- 8% strongly agree AI has transformed how work gets done organizationally
The 57-point gap between “helps me individually” and “has transformed the organization” is structurally consistent with the Brynjolfsson result: AI is generating task-level efficiency, not process redesign. Closing that gap — converting individual task gains into firm-level productivity — is the management challenge of the next phase of enterprise AI adoption.
The Axis Intelligence Research AACI™: AI Assistant Adoption-to-Conversion Index
To quantify the gap between reported AI adoption and demonstrated deep integration, Axis Intelligence Research developed the AI Assistant Adoption-to-Conversion Index (AACI™), a cross-source metric computed separately for consumer and enterprise contexts.
AACI™ Formula:
AACI™ = (Deep Usage Rate ÷ Total Adoption Rate) × (Trust Score ÷ Maximum Trust Score) × 100
Consumer AACI™ (U.S., Q1 2026):
- Deep usage rate: 24% (daily chatbot users, Pew, Feb 2026)
- Total adoption rate: 49% (ever-use chatbot users, Pew, Feb 2026)
- Trust score: 29% (users who trust chatbot output, Pew, Feb 2026; derived from “few say they trust AI to be accurate” per Pew reporting)
- Maximum trust score: 100%
- Calculation: (24 ÷ 49) × (29 ÷ 100) × 100 = 0.490 × 0.29 × 100 = 14.2 / 100
Enterprise AACI™ (U.S., Q1 2026):
- Deep usage rate: 28% (daily or several-times-weekly AI users, Gallup Q1 2026)
- Total adoption rate: 50% (any AI use at work, Gallup Q1 2026)
- Organizational transformation score: 8% (strongly agree AI has transformed work, Gallup Q1 2026 — used as enterprise trust proxy)
- Maximum: 100%
- Calculation: (28 ÷ 50) × (8 ÷ 100) × 100 = 0.56 × 0.08 × 100 = 4.5 / 100
Interpretation: The Consumer AACI™ of 14.2 indicates that less than one in seven users has moved from trial adoption to daily, trusted integration. The Enterprise AACI™ of 4.5 — far lower — reflects that enterprise adoption is largely seat-level rather than workflow-level. Neither score indicates the technology is failing; both scores indicate the sector is earlier in the S-curve than the adoption headlines suggest.
Limitations: The AACI™ is a composite index. Trust score proxies (Pew’s general accuracy trust figure and Gallup’s strong transformation agreement) may understate actual trust in specific task contexts. The index does not distinguish between trust in the tool and trust in AI outputs generally. A higher AACI™ does not imply the tool is being used for high-value tasks. Axis Intelligence Research will re-index this metric quarterly as new Pew and Gallup waves become available.
Public Sentiment on AI Assistants: The Trust Gap
Adoption rising while trust stalls is the defining tension of the AI assistant market in 2026. Pew’s June 2026 data provides the clearest view:
| Sentiment Metric | Value | Source | As-of |
|---|---|---|---|
| Expect AI to benefit society (20-year view) | 16% | Pew | Feb 2026 |
| Expect AI to harm society (20-year view) | 40% | Pew | Feb 2026 |
| Believe AI advancing too quickly | 63% | Pew | Feb 2026 |
| Expect AI to make personal data less secure | 71% | Pew | Feb 2026 |
| No/little confidence in government to regulate AI | 67% | Pew | Feb 2026 |
| No/little confidence in companies to develop AI responsibly | 59% | Pew | Feb 2026 |
| Say chatbots help their productivity | 30% | Pew | Feb 2026 |
| Say chatbots hurt their creativity | 11% | Pew | Feb 2026 |
Source: Pew Research Center, “Americans and AI 2026,” June 17, 2026.
The counterintuitive finding is the age breakdown: adults under 30 are more pessimistic about AI’s societal impact (48% expect harm) than those 50 and older (37% expect harm), despite using chatbots at twice the rate. The heaviest users are also the most skeptical of outcomes. This is not hypocrisy — it reflects that exposure to the technology’s limitations is itself a driver of concern. The users who have experienced hallucinations, bias, and privacy incidents firsthand are the ones revising their long-term outlook downward.
According to Axis Intelligence Research, this pattern — high usage paired with high skepticism — is structurally different from the early internet or early smartphone cycles, where user experience tended to build confidence. It suggests that AI assistant platforms face a trust ceiling that is not simply a matter of time and familiarity, and that improvements in reliability and transparency are commercially necessary, not just ethically desirable.
AI Assistant Adoption by Sector and Region
Sector Variation (Gallup Q1 2026)
Among employees who use AI at work, productivity gains by sector tell a clearer story than headline adoption rates:
| Sector | Extremely Positive Productivity Impact |
|---|---|
| Healthcare | 17% |
| Managerial | 15% |
| Technical and Professional | 13% |
| Office and Administrative Support | 11% |
| Service | 14% (net positive, lower compound effect) |
Source: Gallup Workforce Panel, Q1 2026. gallup.com
Healthcare’s leadership is notable given the sector’s historically slow technology adoption. The combination of documentation burden (AI-assisted note-taking and summarization), decision support, and patient communication creates a task match that is unusually dense relative to other sectors.
Global Context (Stanford AI Index 2026)
The Stanford HAI 2026 AI Index Report — the most rigorously sourced independent data compilation available — provides the global adoption frame:
- Generative AI reached 53% population-level adoption within three years of mass-market launch, faster than the personal computer or the internet at comparable stages
- Global corporate AI investment reached $581.69 billion in 2025, a 129.9% increase from the prior year
- U.S. private AI investment reached $285.9 billion in 2025
- Generative AI accounted for nearly half of all private AI funding in 2025
- AI agent deployment remained in the single digits across nearly all business functions as of early 2026
Source: Stanford HAI, “2026 AI Index Report,” hai.stanford.edu
The U.S. domestic adoption paradox: despite leading the world in AI investment and model development, the U.S. ranks 24th globally in population-level generative AI adoption at 28.3%. Singapore (61%) and the United Arab Emirates (64%) significantly outpace the U.S. per the Stanford Index. The Stanford researchers attribute this to differences in regulatory environment, digital infrastructure maturity, and the composition of the population’s existing digital habits.
Methodology
Data collection: All statistics in this article are drawn from primary sources fetched directly during production in July 2026. The three anchor datasets are:
- Pew Research Center, “Americans and AI 2026: Chatbots, Smart Devices and Views on Impact,” published June 17, 2026. Fielded February 17–23, 2026, via the American Trends Panel probability panel. n = 5,119 U.S. adults. Margin of error ±1.6 pp at 95% confidence. Full report available at pewresearch.org.
- Gallup Workforce Panel, Q1 2026. Published April 13, 2026. Fielded February 4–19, 2026. n = 23,717 U.S. employees working full or part time. Margin of error ±0.9 pp at 95% confidence. Data available at gallup.com.
- Stanford HAI, 2026 AI Index Report. Published 2026. Aggregates data from academic databases, patent filings, government records, and corporate disclosures. Available at hai.stanford.edu.
Market size: Business Research Company’s definition is used consistently to enable time-series comparison. Figures from alternative research houses (MarkWide Research, Technavio, Grand View Research) are noted where referenced but are not primary sourced; their ranges are discussed to illustrate definitional variance, not to report alternative market sizes.
AACI™ methodology: The Axis Intelligence Research AI Assistant Adoption-to-Conversion Index is an Axis-original cross-source metric. Inputs and formula are disclosed inline in the AACI™ section. Inputs: Pew Research (consumer metrics, Feb 2026), Gallup (enterprise metrics, Q1 2026). Limitations are stated in the AACI™ section.
Platform user figures: OpenAI weekly active user figures (900M) come from Reuters citing internal OpenAI data (February 2026). Monthly active user figure (1B) comes from Sensor Tower data cited by Reuters (June 2026). Microsoft Copilot seat figures come from CEO Satya Nadella’s Q3 FY2026 earnings call disclosures (April 29, 2026) as reported by TechCrunch.
Currency: All data in this article reflects sources as of July 2026. The Pew and Gallup surveys reflect February 2026 fieldwork. Stanford data reflects 2025 annual aggregation.
About This Dataset
The accompanying CSV (ai-assistant-statistics-2026.csv) contains 62 rows covering every quantitative claim in this article. Each row includes the value, as-of date, source organization, source document, URL, retrieval date, and whether the figure is primary-sourced or Axis-calculated. The AACI™ rows carry full formula disclosure in the method_note column.
License: CC BY 4.0. You may use, share, and adapt this dataset with attribution.
Citation: Axis Intelligence Research, “AI Assistant Statistics 2026,” Axis Intelligence Research, 2026. https://axis-intelligence.com/ai-assistant-statistics/
Cite This Article
APA Axis Intelligence Research. (2026). AI assistant statistics 2026: Adoption, usage, and the productivity gap. https://axis-intelligence.com/ai-assistant-statistics/
MLA Axis Intelligence Research. “AI Assistant Statistics 2026: Adoption, Usage, and the Productivity Gap.” Axis Intelligence Research, 2026, axis-intelligence.com/ai-assistant-statistics/.
Chicago Axis Intelligence Research. “AI Assistant Statistics 2026: Adoption, Usage, and the Productivity Gap.” Axis Intelligence Research. 2026. https://axis-intelligence.com/ai-assistant-statistics/.
FAQ
What percentage of people use AI assistants in 2026?
In the U.S., 49% of adults have used an AI chatbot as of early 2026, up from 33% in 2024 (Pew Research Center, February 2026). Daily use stands at 24%. Globally, generative AI has reached 53% population-level adoption within three years of mass-market launch, outpacing the spread of both the personal computer and the internet at comparable stages (Stanford HAI 2026 AI Index).
Which AI assistant has the most users?
ChatGPT leads by a wide margin. According to OpenAI (via Reuters, February 2026), ChatGPT reached 900 million weekly active users globally. In the U.S. adult population, 44% report using ChatGPT — more than Gemini (24%), Microsoft Copilot (17%), and Meta AI (14%) combined on a single-platform basis (Pew Research Center, June 2026).
How many enterprise users does Microsoft Copilot have?
Microsoft 365 Copilot crossed 20 million paid enterprise seats as of April 29, 2026, per CEO Satya Nadella’s Q3 FY2026 earnings disclosure. Those seats represent approximately 4.4% of the 450+ million commercial Microsoft 365 subscribers — meaning the majority of Microsoft’s potential Copilot market has yet to adopt the product.
Do AI assistants actually improve productivity?
Yes, at the task level — with significant variation by worker type. The most rigorous available RCT (Brynjolfsson, Li, and Raymond, NBER/QJE 2025) found a 14% average productivity gain for customer support agents using AI assistance, with 34% gains for novice workers and near-zero gains for top performers. At the organizational level, Gallup Q1 2026 finds only 8% of employees in AI-adopting organizations strongly agree that AI has transformed how work gets done. The gap between task-level gain and organizational transformation is the central finding of enterprise AI research in 2026.
How much is the AI assistant market worth in 2026?
The global virtual assistant market is projected to reach $10.11 billion in 2026, up from $8.11 billion in 2025, at a 24.7% CAGR (Business Research Company, 2026). The estimated U.S. consumer surplus from generative AI tools alone reached $172 billion annually by early 2026 (Stanford HAI 2026 AI Index) — a figure that dwarfs commercial revenue because most leading tools remain free or near-free.
Why do people not trust AI assistants even as they use them?
Pew Research Center’s June 2026 survey finds that 71% of U.S. adults expect AI to make their personal information less secure, 63% believe AI is advancing too quickly, and only 16% expect AI to benefit society over 20 years — despite 49% using chatbots and 24% using them daily. The heaviest users (adults under 30) are also the most pessimistic about societal impact (48% expect harm). Direct exposure to AI limitations — hallucinations, privacy incidents, and output inconsistency — appears to drive skepticism alongside adoption, not away from it.
What are AI assistants most commonly used for?
Information search is the top use case: 42% of U.S. chatbot users use them to find information. Work tasks (38% of employed users), entertainment (25%), image or video creation (24%), and medical advice (20%) follow. Emotional support (10%) and companionship (4%) are smaller but fastest-growing categories relative to their base (Pew Research Center, June 2026).
How does AI assistant adoption differ by age?
Adults under 50 use AI chatbots at roughly twice the rate of those 50 and older. Among 18–29-year-olds, usage reaches 66%. Among adults 65 and older, approximately 75% report never using a chatbot (Pew Research Center, June 2026). ChatGPT specifically: adults under 50 use it at 57%, compared to 28% among those 50 and older.
What is the Axis AACI™ score for AI assistants?
Axis Intelligence Research calculated the AI Assistant Adoption-to-Conversion Index (AACI™) at 14.2/100 for U.S. consumers and 4.5/100 for U.S. enterprise environments as of Q1 2026. The index measures the ratio of daily/deep users to total adopters, weighted by a trust or transformation score. A score of 14.2 indicates that roughly one in seven consumer adopters has achieved genuine daily integration paired with trust in outputs. The enterprise score of 4.5 reflects that seat deployment has far outpaced workflow-level transformation.
How fast has ChatGPT grown?
ChatGPT’s growth is the fastest adoption curve in consumer technology history by available measures. It reached 1 million users in 5 days of launch (November 2022), 100 million monthly active users in 60 days (January 2023, Reuters), 400 million weekly active users in February 2025 (OpenAI), 900 million weekly active users in February 2026 (OpenAI via Reuters), and 1 billion monthly active app users in June 2026 (Sensor Tower via Reuters). That trajectory — from zero to 900 million weekly users in roughly 40 months — has no precedent in consumer technology.