18 pts
Gap between US adults (48%) and AI experts (30%) expecting AI to harm decision-making and problem-solving by 2035, the widest of 12 capacities
What the chart shows
In Elon University’s 2025 survey, US adults were more likely than AI experts to expect AI to do more harm than good by 2035 to 9 of 12 human capacities. The largest gaps are for decision-making and problem-solving (48% vs 30%), metacognition, the ability to think about one’s own thinking (53% vs 36%), and innovative thinking and creativity (45% vs 30%).
Social and emotional intelligence draws the most concern from both groups: 55% of adults and 50% of experts. Experts are more negative than adults on three: trust in shared values and norms (48% vs 41%), confidence in one’s own abilities (48% vs 43%) and mental well-being (45% vs 42%).
Key numbers
| Measure | Value |
|---|---|
| Decision-making and problem-solving, US adults | 48% |
| Decision-making and problem-solving, AI experts | 30% |
| Metacognition, US adults | 53% |
| Metacognition, AI experts | 36% |
| Social and emotional intelligence, US adults | 55% |
| Social and emotional intelligence, AI experts | 50% |
How to read it
- Elon University’s Imagining the Digital Future Center surveyed 1,005 US adults on 17–20 July 2025 and compares them with experts; the AI Index labels that group AI experts.
- The AI Index text gives 51% (adults) and 34% (experts) for social and emotional intelligence, and 55% and 53% for mental well-being. The data file and the report’s chart show 55% and 50%, and 42% and 45%, which we use.
Source: Elon University, 2025, via the AI Index Report 2026 (Stanford HAI), figure 9.2.2: Impact of AI on key human capacities and traits: general public vs. experts. Chart and text by AI Adoption.
Numbers checked against the published data on 20 October 2026.
Download the data (CSV) · Cite this chart
Numbers checked against the published data on 20 October 2026.
Download the data (CSV) · Cite this chart
Discussion