Anthropic Releases Large-Scale AI Survey of 80,000 People: 81% of Users Say AI Has Delivered on Its Promises, Western Europe Most Pessimistic, Africa and Latin America Most Optimistic

According to CoinWorld, based on monitoring by 1M AI News, Anthropic released its largest AI qualitative research report to date yesterday. In December last year, the team invited Claude users to participate in one-on-one interviews with AI interviewers, collecting 80,508 interviews from 159 countries in 70 languages. Anthropic claims this is the largest and most comprehensive qualitative study in history. Users’ visions for AI are divided into nine categories, with the highest percentage (18.8%) hoping AI will handle chores to focus on high-value work, followed by “personal growth” (13.7%) and “life management” (13.5%). 81% of respondents said AI has, to some extent, progressed toward their described vision, with time savings being the most frequently mentioned realized benefit (50%). Regarding concerns, unreliability (hallucinations, citation errors, etc.) tops at 26.7%, employment and economic impacts are second at 22.3%, and worries about loss of human autonomy are third at 21.9%. Concerns about employment and the economy are the strongest predictive factors affecting overall AI sentiment.

67% of global respondents hold a positive attitude toward AI, but regional differences are significant. Latin America (Peru 82%) and Sub-Saharan Africa (Nigeria 81%) are the most optimistic, viewing AI as a lever to overcome barriers in capital and education; Western Europe (UK 63%), North America (USA 66%), and Oceania tend to be more negative, highly correlated with greater concerns about AI’s economic impacts in these regions. East Asia’s concern structure is unique, with lower worries about governance and monitoring than the global average, but more prominent concerns about cognitive decline (18%) and loss of meaning (13%). “The West worries about who controls AI, while East Asia worries more about the personal impact of using AI.”

The core concept of the report is “Light and Shadow”: the same AI capability can bring both benefits and risks, often coexisting within the same individual. Those expecting emotional companionship from AI are three times more likely to worry about becoming dependent on it.