AI Statistics
19 figures on AI adoption, productivity and reliability. Each one links to its primary source.
We include a figure only when we opened the original report or study and read the number there. Each figure shows the period it covers. All figures were last checked on .
Statistics by topic
- AI adoption statisticsHow many organizations, workers and adults use AI, with a primary source for each figure.
- AI productivity statisticsWhat controlled studies and national surveys measured when people used generative AI for real work.
- AI reliability and risk statisticsHow often AI gets real tasks right, where it still fails and how reported incidents have grown.
Headline figures
- 88% of organizations use AI, according to the Stanford 2026 AI Index.
- Generative AI users reported saving 5.4% of their work hours, about 2.2 hours in a 40-hour week, according to the St. Louis Fed.
- AI agents completed 66.3% of computer tasks on the OSWorld test in 2025. The human baseline is 72.35%, according to the Stanford 2026 AI Index.
How we choose figures
- We use primary sources: the research group, the survey firm or the journal that produced the number.
- We state what the source measured, because two surveys of AI use often ask different questions.
- We keep the period next to each figure, so you can tell a 2023 experiment from a 2026 survey.
- The same data is available as one file at /statistics/ai-statistics.json.
For plain definitions of the terms in these pages, see the AI Encyclopedia. To estimate your own time savings, use the AI time savings calculator.