Anthropic Asked 81,000 People What They Want From AI. Here's What They Said.

Anthropic's global AI study surveyed 80,508 participants across 159 countries, revealing desires for more personal time and concerns about AI's unreliability and job displacement. Sentiments vary regionally, with lower-income countries seeing AI as an equalizer, while Western Europe and North America focus on governance issues. The study highlights a complex mix of hope and fear regarding AI's impact.

Anthropic just published what it claims is the largest qualitative AI study ever conducted - and the results are more nuanced than the usual optimist/pessimist split you see in tech media.

The study

Between December 2024 and January 2025, Anthropic used a version of Claude as an interviewer, inviting all Claude.ai users to share how they use AI, what they hope for, and what scares them. 80,508 people across 159 countries and 70 languages responded. That's not a survey with checkboxes - it's open-ended interviews, analyzed at scale using Claude-powered classifiers.

What people actually want

The #1 ask wasn't "help me be more productive." The dominant theme was simpler: more time back for the things that matter outside of work.

• 19% want AI to help them achieve professional excellence - less busywork, more meaningful work
• 14% want personal transformation - emotional growth, mental health support, companionship
• 11% explicitly want AI to free up time for family, hobbies, and rest
• 10% are chasing financial independence through AI

A common pattern: people said they wanted AI for email automation or documentation - but when asked why, the real answer was "so I can pick up my kids on time."

The fears

Concerns were more varied than hopes, and most people voiced 2.3 distinct worries on average. The top concern was unreliability and hallucinations (27%), followed closely by job displacement and loss of human autonomy, both at 22%. Cognitive atrophy came next at 16%, then governance gaps at 15%, and misinformation at 14%.

Only 11% said they had no concerns at all.

The most interesting finding

Hope and fear don't divide people into separate camps - they coexist in the same person. Someone excited about AI for emotional support is 3x more likely to also fear becoming dependent on it. The people most enthusiastic about AI-powered learning are among the most worried about cognitive atrophy.

The exception: economic anxiety. The people excited about AI-driven entrepreneurship and the people worried about displacement are largely different groups - not the same person holding both at once.

The geography gap

Sentiment toward AI skews heavily by region. Lower and middle income countries are consistently more optimistic - not because they're naive, but because AI registers more as an equalizer than a threat.

"I'm in a tech-disadvantaged country, and I can't afford many failures. With AI, I've reached professional level in cybersecurity, UX design, marketing, and project management simultaneously." - Entrepreneur, Cameroon

Western Europe and North America cluster on the opposite end: higher concern about governance, surveillance, and job displacement. 67% of all respondents expressed net positive sentiment toward AI - but that number shifts significantly depending on where you live.

Bottom line

This is a rare piece of qualitative research that cuts through the usual AI discourse. The findings aren't about abstract risk projections, they're grounded in what 80,000 actual users are experiencing day to day. For anyone building or governing AI products, it's required reading.

• https://www.anthropic.com/features/81k-interviews...