AI for Automation
Back to AI News
2026-03-19AnthropicAI surveyClaudeAI adoption

What 81,000 people actually want from AI

Anthropic surveyed 81,000 Claude users across 159 countries. The #1 hope: automating boring work. The #1 fear: AI making things up.


Anthropic just published the largest survey ever conducted about what people actually think about AI. They interviewed 80,508 Claude users across 159 countries in 70 languages — and the results are surprisingly honest about both the excitement and the anxiety.

The headline finding: 67% of users worldwide feel positive about AI, but nearly everyone has at least two distinct worries about where it's heading. The average person flagged 2.3 separate concerns.

What 81,000 people want from AI - Anthropic study

The 9 Things People Want Most

When asked to describe their ideal AI future, respondents clustered around these visions — ranked by how many people picked each one:

1. Automate the boring stuff (18.8%) — The top answer by far. People want AI to handle routine tasks so they can focus on meaningful, strategic work.

2. Personal growth & emotional support (13.7%) — Surprisingly, the second-most-wanted use is therapy-like support and companionship.

3. Life management (13.5%) — Help organizing schedules, reducing mental load, and managing the chaos of daily life.

4. More free time (11.1%) — Reclaiming time for family, hobbies, and rest.

5. Financial independence (9.7%) — Using AI to generate income or build economic security.

Rounding out the list: societal change (9.4%), entrepreneurship (8.7%), learning (8.4%), and creative expression (5.6%). The pattern is clear — people don't just want AI to do tasks. They want it to give them their lives back.

What AI Has Actually Delivered

The good news: 81% of respondents said AI has already helped them. Here's how:

Productivity boost (32%) — The most common experience. People reported dramatically faster work and less time on repetitive tasks.
Thinking partner (17.2%) — People use AI as a brainstorming collaborator and sounding board for ideas.
Built things they couldn't before (8.7%) — Non-technical users creating apps, websites, and projects that used to require a developer.

One user quote stood out: "Claude put historical pieces together, leading to my proper diagnosis after 9 years of misdiagnosis." Another: "I can now leave work on time to pick up my kids, feed them, and play with them."

But 18.9% said AI fell short of their expectations — it wasn't capable enough or produced unreliable results.

The 5 Biggest Fears

Here's what keeps AI users up at night:

1. AI makes things up (26.7%) — Hallucinations (when AI confidently states false information), inaccuracies, and fake citations topped every other concern. More than one in four users flagged this.
2. Job loss (22.3%) — Nearly a quarter worry about displacement, unemployment, and wage stagnation.
3. Loss of human control (21.9%) — People fear being forced to adopt AI or losing autonomy over their decisions.
4. Brain rot (16.3%) — Worry about losing the ability to think, learn, and solve problems independently. One lawyer from Israel put it bluntly: "I fear losing my ability to read by myself. Thinking was the last frontier."
5. No real oversight (14.7%) — Lack of laws, regulations, and governance frameworks.

A fascinating finding: educators noticed "brain rot" 2.5 to 3 times more than the average respondent. And lawyers showed the highest rates of both benefiting from AI decisions (54%) and being harmed by unreliable outputs (46%).

The World Doesn't Agree

The survey revealed striking regional divides:

Most optimistic: Sub-Saharan Africa (76% positive), South Asia, and Latin America — regions where AI represents opportunity and upward mobility. Entrepreneurship was the dominant theme.

Most skeptical: Western Europe (64% positive) and North America (66%) — wealthier regions where people worry more about job loss and losing control.

Most anxious about thinking skills: East Asia, where 18% expressed concern about cognitive atrophy (compared to 16% globally) and 13% worried about loss of meaning.

Independent workers and freelancers experienced economic gains from AI 3 times more often than people working at large companies (47% vs. 14%).

The Paradox at the Heart of AI

The study's most striking insight: the same AI capabilities that help people are the ones that scare them. People who value AI for emotional support are 3x more likely to fear becoming dependent on it. Half of users say AI saves them time, but 19% say it actually creates more work because they spend so long verifying outputs.

It's not that some people love AI and others hate it. Most people feel both — at the same time, about the same features.

You can read the full study at Anthropic's website.

Related ContentGet Started with Easy Claude Code | Free Learning Guides | More AI News

Stay updated on AI news

Simple explanations of the latest AI developments