News Analysis
Anthropic has released its third Economic Index report, providing one of the most detailed snapshots yet of how artificial intelligence — specifically its Claude model — is being adopted across states, countries, industries and businesses.
The findings suggest AI is becoming more prevalent and increasingly trusted to carry out complex work with little human oversight.
Uneven But Rapid AI Adoption
The report shows that AI adoption is remarkably uneven, both globally and within the United States. At the international level, the US remains the dominant user of Claude, accounting for more than 21% of global usage. India, Brazil, Japan and South Korea round out the top five.
But raw usage numbers don’t tell the full story. Anthropic created the Anthropic AI Usage Index (AUI) to measure adoption relative to the working-age population. By that metric, smaller, high-income nations like Israel, Singapore, Australia, New Zealand and South Korea lead the way. The data reveals a strong correlation between GDP per capita and AI adoption: for every 1% increase in GDP per capita, AUI rises by 0.7%.
That link raises an important concern. Historically, transformative technologies like electrification and the combustion engine widened global inequality, with richer countries accelerating ahead. AI could follow a similar path if its benefits concentrate in wealthier economies with robust internet infrastructure and knowledge-based industries.
Related Article: When AI Learns to Hack: Inside CMU and Anthropic’s Cybersecurity Experiment
AI Patterns Inside the United States
The same income-based correlation appears when comparing US states — but the picture is more complicated. A 1% increase in per capita GDP is linked with a 1.8% increase in Claude use, but income alone doesn’t explain everything.
Instead, local industry composition plays a defining role. For example:
- District of Columbia tops the US AUI at 3.82, with heavy use in editing, information retrieval and document-related tasks common to government and policy work.
- California sees outsized use in coding-related tasks, consistent with its technology-heavy economy.
- New York ranks fourth overall, with Claude heavily used for finance-related tasks.
- Hawaii uses Claude disproportionately for tourism-related work, reflecting its hospitality-driven economy.
AI is embedding itself not just broadly, but specifically within the economic DNA of each region.
From Augmentation to Automation
Perhaps the most striking shift revealed in the report is the growing reliance on automation. Anthropic tracks how people use Claude by categorizing interactions into augmentation (collaborative tasks) and automation (AI completing work with little oversight).
Back in December 2024, directive automation accounted for 27% of usage. By mid-2025, that number had jumped to 39%, tipping the overall balance in favor of automation (49.1%) over augmentation (47%).
The change suggests growing trust in AI’s capabilities. As models like Claude improve at anticipating user needs and generating high-quality outputs, more people are willing to accept results on the first attempt rather than iterate collaboratively.
Still, the global pattern is nuanced. In countries with higher per-capita use, people lean toward augmentation, while lower-use countries rely more heavily on automation. That may reflect cultural differences, the maturity of adoption or resource constraints that encourage fully outsourcing tasks to AI.
Changing Nature of Work
The types of work people delegate to Claude are also shifting. While software engineering remains the dominant use worldwide, other sectors are catching up fast. Since December 2024:
- Educational tasks have risen more than 40%, from 9% to 13% of all conversations.
- Physical and social sciences increased by one-third, from 6% to 8%.
- Business and financial operations declined from 6% to 3%.
- Management-related tasks dropped from 5% to 3%.
As adoption spreads into knowledge-intensive fields, Claude is being asked to do more diverse work, moving beyond coding into education, research and content-heavy domains.
Related Article: Inside Anthropic’s $1.5B Generative AI Copyright Lawsuit Settlement
Businesses Lean Heavily on Automation
For the first time, Anthropic included anonymized data from its API customers — businesses and developers that integrate Claude directly into their systems. Their usage patterns look very different from individual consumers.
- 44% of API traffic is tied to computer or mathematical tasks, compared to 36% on Claude.ai.
- Automation dominates business use: 77% of API conversations are automated, with only 12% showing augmentation.
- Businesses are more likely to pay for higher-cost, token-intensive tasks.
If these patterns hold, the economic implications could be significant. Heavy reliance on AI automation within enterprises suggests the potential for major productivity gains, and possible labor market disruption, as companies entrust more responsibility to machines.
Why It Matters
Anthropic’s Economic Index is still in its early stages, but it offers one of the clearest windows yet into how AI is weaving itself into daily life and work.
Webinar
Sep
25
CMS Briefing: A Live Look at What’s Next in AI-Driven Platforms
Learn how leading organizations are using AI‑driven tools to publish faster, personalize smarter and stay secure.
Register
Webinar
Oct
15
Agentic AI Playbook Real-World Customer Service Use Cases You Can Deploy Now
Boost self-service by 30% and slash call volume by 63% with agentic AI.
Register
Webinar
On demand
Ready or Not: How Data-First Organizations Are Unlocking Agentforce Potential
Learn how to cut through the noise, activate Agentforce and build a Salesforce AI strategy that actually delivers.
Watch Now
Webinar
On demand
AI in Customer Service: Faster Resolutions, Happier Customers
Don’t let rising demand burn out your team. See how to build a smarter, more resilient support org.
Watch Now
Webinar
On demand
From Hype to High-Impact CX Strategies That Actually Scale
Turn buzzworthy AI and outsourcing trends into measurable CX wins with fresh 2025 data.
Watch Now
Webinar
On demand
Insights to Action Rethinking the Contact Center for Real Business Impact
Join our exclusive webinar to hear CX executives share their innovative strategies for transforming service delivery.
Watch Now
On one hand, organizations and workers who embrace AI may benefit from efficiency and competitive advantages. On the other, the concentration of adoption in wealthy economies and industries could deepen economic divides, both globally and domestically.
What’s clear is that people are becoming increasingly comfortable letting AI do the work. Whether editing research papers in Massachusetts, planning travel in Hawaii or coding applications in India, Claude is steadily becoming a co-worker, and in many cases, an autonomous one.
As Anthropic put it, “The nature of people’s use of Claude is evidently still being defined: we’re still collectively deciding how much confidence we have in AI tools, and how much responsibility we should give them.”
FAQs
While software engineering remains the most common task globally, other areas are growing rapidly:
- Educational tasks rose from 9% to 13% of all conversations.
- Physical and social sciences grew from 6% to 8%.
- Business and financial operations fell from 6% to 3%.
- Management tasks dropped from 5% to 3%.
Anthropic’s API customers — mostly businesses and developers — use Claude far more for automation. 77% of business interactions are automated, compared to nearly even automation-augmentation usage on Claude.ai. Businesses also concentrate on coding and administrative tasks to achieve potential productivity gains and labor market impacts.
Yes. Historically, transformative technologies like electrification and the combustion engine spurred growth but also widened inequality. If AI adoption continues to concentrate in wealthy nations and knowledge economies, it may deepen global and regional economic divides.
Anthropic uses a privacy-preserving classification method to analyze anonymized conversation transcripts. These are categorized using O*NET, a US government database of occupations and tasks. This data allows researchers to track patterns in how Claude is used over time without exposing individual users’ information.
Anthropic has created an interactive Economic Index website where users can explore AI adoption trends by state, country and occupation. The company has also made its dataset openly available for policymakers, researchers and the public.
