# What 81,000 people told Anthropic about AI

Author: Norbert Biedrzycki
Published: 2026-03-24
Language: en
URL: https://norbertbiedrzycki.pl/en/what-81000-people-told-anthropic-about-ai/
Summary: Anthropic spent a week interviewing 81,000 people about their relationship with AI. Not a survey.

Anthropic spent a week interviewing 81,000 people about their relationship with AI. Not a survey. Actual conversations – in 70 languages, across 159 countries.

A few things I can't stop thinking about after reading it.

Someone is using AI at 3am because they're having a hard night and their therapist isn't available. They're not replacing human connection, but they're getting through until they can find it. 6% of users named emotional support as something AI has genuinely given them. That same group is three times more likely to also fear becoming dependent on it. Same person. Same time. Both things true.

Teachers are watching cognitive atrophy happen in real time in their students. They report it at 2–3x the rate of average users. That's not a hypothetical future risk. It's a Tuesday.

And the global split surprised me. In Europe and the US, people want AI to take the edge off: the endless admin, the mental load, the feeling of drowning in complexity. In Sub-Saharan Africa and South Asia, the framing is completely different: AI as a way to start a business without needing capital. Same tool, completely different world.

The honest takeaway from the whole report: people aren't optimists or pessimists about AI. They're both, at the same time, about the same thing. The thing that makes AI feel supportive is exactly what makes it easy to become dependent on. You can't fix that with a guardrail.

That's the actual design challenge.
