Anthropic survey: expectations and concerns about artificial intelligence

A representative Anthropic survey of almost 52,000 people in the United States paints a mixed picture of expectations surrounding artificial intelligence. The greatest hope is for medical progress: 48% of respondents would like AI to help cure serious diseases such as cancer or Alzheimer’s in the future. 36% hope it will provide better support for people with disabilities, while only 23% mention making everyday life easier through technological advances. Respondents were far less likely to associate AI with hopes for therapeutic support or reduced loneliness.

Concerns are focused primarily on the impact on jobs. 64% fear that AI could put employment at risk; this is the most common concern in every US state and becomes more pronounced as education levels rise. 56% also see a risk of becoming cognitively dependent on AI or thinking less independently, while 52% cite misinformation as an additional concern.

Overall, attitudes towards the use of AI in the workplace remain sceptical. Acceptance increases, however, when AI is perceived as competent in a particular area. People who regularly use AI at work also assess the risk of losing their own job as lower than those who do not use AI in their day-to-day work at all.

By Ida Kandler

Source: The Decoder, 12 June 2026, by Matthias Bastian