Anthropic just published the results of one of the largest and most personal AI studies ever conducted. For roughly a month, Claude users around the world were invited to have an open-ended conversation about their hopes and fears for artificial intelligence. No fixed questionnaire — every dialogue adapted in real time based on what the user said. Then Claude’s own classifiers analyzed the responses.
80,508 people from 159 countries answered in 70 languages.
Here’s what emerged when the numbers were crunched.
What People Actually Want from AI
The top aspirations (as a percentage of all responses):
- Professional efficiency — 18.8%;
- Personal transformation — 13.7%;
- Life management & cognitive offloading (schedules, reminders, mental load) — 13.5%;
- More free time for family and self — 11.1%;
- Financial independence — 9.7%;
- Societal transformation (medicine, climate, inequality) — 9.4%;
- Entrepreneurship — 8.7%;
- Learning & personal development — 8.4%;
- Creative self-expression — 5.6%.
Strikingly, 81% of respondents said AI has already taken at least one meaningful step toward their personal dream.
What People Fear Most
The biggest concerns (again as percentages):
- Unreliability & hallucinations — 26.7%
- Job loss — 22.3%
- Loss of autonomy & control — 21.9%
- Cognitive atrophy (“my brain will get lazy”) — 16.3%
- Lack of regulation — 14.7%
- Misinformation & deepfakes — 13.6%
On average, each person expressed 2.3 distinct fears. Only 11% said they see no risks at all.
The Real Insight: Hope and Fear Are Inseparable
Anthropic calls these “tensions” — moments where the very thing people love about AI is also what terrifies them.
Example: People who value AI most for emotional support are three times more likely to fear emotional dependency. Lawyers report the highest real-world gains from AI in decision-making — and also the highest frustration with hallucinations. The same pattern repeats across professions and life goals: the upside and downside are two sides of the same coin.
Geography Tells a Clear Story
Overall, 67% of respondents feel positive about AI’s future.
But the divide is stark:
- Most optimistic regions: Africa, Latin America, and South Asia — where AI is often seen as a social elevator.
- Most concerned regions: Europe and North America — where the dominant fear is cognitive overload and job displacement.
Fear of losing one’s job is by far the strongest predictor of negative sentiment toward AI.
Also read:
- The "Make-Your-Own" Apocalypse: Why Democratized Coding Could Liquidate the Marketing Industry
- American Marketing Trends Set to Dominate 2026: A Global Ripple Effect
- The Perils of AI Companions: FoloToy's Kumma Bear and the Dark Side of Smart Toys
Bottom Line
This wasn’t just another poll. It was a global, deeply personal conversation with more than 80,000 humans who talked directly to an AI about what they want it to become — and what they’re scared it might turn into.
The data shows we don’t have clean “pro-AI” and “anti-AI” camps. Most of us are both at the same time: excited by the promise and uneasy about the price.
You can read the full report with detailed quotes, breakdowns by country, profession, and age here:
anthropic.com/features/81k-interviews
The age of AI is no longer coming.
It’s already here — and 80,508 real conversations just told us exactly how we feel about it.
