Optimism.
With receipts.

90% reported saving time. 84% reported more creativity. 83% reported enjoying work more.

Who and how. Edelman Data & Intelligence surveyed 31,000 full-time employed or self-employed knowledge workers across 31 markets, February 15–March 28, 2024. The percentages featured here describe respondents who use AI. The research is sponsored by Microsoft and LinkedIn.

What it doesn’t tell us. These are self-reported benefits among AI users, not measured productivity gains or effects for all workers.

U.S. public K–12 teachers who used AI at least weekly estimated saving an average of 5.9 hours each week.

Who and how. The study surveyed 2,232 U.S. public-school teachers recruited from RAND’s American Teacher Panel, March 18–April 11, 2025. Teachers estimated time changes in half-hour increments across nonoverlapping tasks. Qualitative responses described individualized lessons, student feedback, and family time.

What it doesn’t tell us. The time-savings figure applies to weekly AI users. It is a self-estimate, not a measured causal effect or an estimate for every teacher.

One in four workers worldwide is in an occupation with some potential exposure to generative AI. Job transformation is the most likely overall impact.

Who and how. The refined index combines task assessments, worker and expert input, AI scoring, and employment data. Most occupations include tasks that continue to require human input.

What it doesn’t tell us. Exposure describes potential changes to tasks. It does not measure actual adoption, displacement, or job creation, and it does not rule out job losses.