Anthropic data shows uneven AI adoption
Add Axios as your preferred source to
see more of our stories on Google.
Illustration: Maura Kearns/Axios
AI use is spreading unevenly, with wealthier countries, people with more skills and certain U.S. states more likely to benefit, according to new usage data from Anthropic.
Why it matters: The data shows workers adopting AI faster than past tech advances, but uneven adoption threatens to worsen existing inequalities.
The big picture: The interactive report breaks down how usage of the Claude chatbot varies by country as well as by state — within the U.S. — identifying which states are using the technology more and for what purposes.
Zoom in: California, Washington, D.C., and Utah show the largest usage relative to population, while a number of southern and Plains states are underrepresented.
- People are also using AI for different tasks across states.
- In Hawaii, tourism is a major driver of Claude usage, while in D.C. a lot of people are using the chatbot for job searches and writing tasks.
- Coding is a major task for people in California.
- Higher-skilled workers are benefiting more than entry-level workers, perhaps because they can fill in gaps in Claude's knowledge.
Zoom out: Globally, wealthier countries — including Israel, Singapore and Canada — have greater use per capita, while emerging countries such as India and Nigeria lag in usage.
- In wealthier countries with broad access, Claude is used mainly to augment workers. In lower-income countries with less access, it's used more for automation.
- However, researchers cautioned it was too early to predict what implications that might have for long-term job displacement.
Between the lines: Anthropic's report highlights the risks of AI repeating patterns seen in past tech waves — improving living conditions and economic growth broadly, but with benefits accruing mostly to the wealthy.
- "If the productivity gains are larger for high-adoption economies, current usage patterns suggest that the benefits of AI may concentrate in already-rich regions — possibly increasing global economic inequality and reversing growth convergence seen in recent decades," Anthropic said in the report.
What they're saying: Anthropic economist Peter McCrory told Axios that the goal of the research is to equip policymakers, researchers and the broader public with enough data "so that the benefits of AI are broadly felt."
- "What we really hope with this data is to help people understand [and] anticipate how AI at the frontier is changing the nature of work," McCory said.
- Anthropic is making the data openly available for other researchers and policymakers.
- "This is one way we live that value of transparency," Anthropic external affairs head Sarah Heck told Axios.
The fine print: The company collected a random sample of about 1 million conversations from early August 2025.
- It then used a tool called Clio ("Claude insights and observations"), which tracks Claude usage while preserving user privacy.
- Anthropic also mapped conversations to a Labor Department taxonomy of occupations and tasks to infer job usage.
- Geographic data was inferred by IP address.
What's next: In addition to making this round of data broadly available, Anthropic said it's committed to future research to study how usage changes as AI adoption grows.
- "The AI usage index introduced in the report will be a very useful framework for understanding how this diffuses throughout the economy," McCory said. "I'm very excited to track that over time."
