Anthropic has updated its "Anthropic Economic Index," an interactive dashboard built on 1 million real conversations from Claude.ai and its API, spanning 116 countries and 51 U.S. states. The data reveals that 49% of global occupations now have at least 25% of their tasks covered by AI, up from 36% in January 2025—a 13 percentage point jump in just over a year.
Collaboration vs. Automation: The Balance Shifts
Currently, augmentation (human-AI collaboration) accounts for 52% of usage, while full automation makes up 45%. Back in August 2025, automation held a slight lead at 49% versus 47%. This suggests AI is moving from "taking your job" to "working alongside you," at least for now.
By use case, coding represents 35% of all Claude conversations, the single largest category. Experienced users show a clear edge: those who have used Claude for more than six months achieve a 10% higher task success rate and use it for work-related purposes 7% more often than newcomers.
Taiwan Ranks 39th: Tech Troubleshooting and Academic Tasks Dominate
Among 116 countries, Taiwan ranks 39th with a usage index of 1.66, meaning its per capita Claude usage is 66% above the global expected value, based on 5,930 conversations. Top use cases include resolving hardware/software/system issues (5.1%), assisting with academic assignments across subjects (4.1%), developing/debugging websites and web applications (3.3%), and programming/software development/multilingual coding (2.6%).
By occupation, computer and mathematics fields account for 24.4% of all usage, far outpacing office and administrative support (5.9%), arts and design (5.7%), education and instruction (5.2%), and life and social sciences (1.6%).
Check Your Own Occupation Now
The Anthropic Economic Index is not just a report—it's an interactive query tool. Users can visit the homepage, enter an occupation name (e.g., "Software Developer" or "Financial Analyst"), and instantly see the proportion of tasks already covered by AI, broken down into augmentation and full automation. It offers a concrete baseline for assessing AI's impact on specific roles, going beyond mere intuition.