Economic Index Primitives

The Anthropic Economic Index report: New building blocks for understanding AI use \ Anthropic Economic Research Anthropic Economic Index: New building blocks for understanding AI use Jan 15, 2026 Read the full report

Is artificial intelligence really making people faster at work? What sort of tasks does AI support best? And how might it change the nature of people’s occupations? At Anthropic, we’re measuring real-world AI use on an ongoing basis to answer questions exactly like these. Our privacy-preserving analysis method allows us to learn more about conversations on Claude.ai (capturing uses by consumers) and our first-party API (mostly capturing uses by businesses). 1 In past reports, we’ve assessed AI tasks by occupation and wage level , looked more closely at software development , and studied AI use by country and by US state . We’re now adding a new level of detail to our Economic Index. In our fourth report, we’re introducing what we’ve called economic primitives : a set of five simple, foundational measurements to track the economic impacts of Claude over time. Our initial set includes task complexity, skill level, purpose (work, education, or personal use), AI autonomy, and success. 2 We derive these primitives from asking Claude to answer a common set of questions about every conversation in our sample for this report. These primitives provide a leading indicator of AI’s potential economic impacts—and allow us to answer far more complex questions about how AI is already changing jobs. Our latest report, which samples conversations from November 2025 (predominantly using Claude Sonnet 4.5), uses our primitives to explore a wide range of questions that we wouldn’t otherwise be able to answer—including how Claude’s task-level success rates change for more complex tasks, and whether the use of Claude to date might portend a net-deskilling effect on many jobs. You can read the fourth Economic Index report here . Below, we summarize its results. What we’ve learned from our economic primitives We applied our economic primitives to questions about individual tasks, occupations, and then the possible aggregate impacts of the changes we observe. (Our full methodology—including details on how we tested the accuracy of our primitives—is described in chapter two of the full report .) Tasks Which tasks does AI speed up, and by how much? We found that more complex tasks were sped up the most by Claude. We measure this by what Claude estimates as the number of years of schooling required to understand the conversation’s inputs: in Claude.ai, tasks with prompts requiring a high school education (12 years) were sped up by a factor of 9, while those requiring a college degree (16 years) were sped up by a factor of 12. (On the API, the speedup was greater still.) These results imply that AI’s productivity gains are currently accruing in tasks that require relatively high human capital, which is consistent with the evidence that white collar professionals are more likely to use AI at work. This same trend holds—albeit in weaker form—when we adjust for tasks’ success rates. Claude successfully completes tasks that require a college degree 66% of the time, compared to 70% for those tasks that require less than a high school education. This reduces, but doesn’t eliminate, the overall effect: Claude’s impact on task speedup scales more sharply with complexity than complexity correlates with a decrease in success rate. Speedup and success rate vs. human years of schooling. The chart on the left shows a scatterplot of the relationship between speedup and human years of schooling, measured at the O*NET task level. The dashed lines show the line of best fit. The chart on the right shows the relationship with the success rate. What are the time horizons over which Claude can support tasks? METR ’s measure of AI’s task horizons shows that longer tasks are harder for AI models to complete. But the length of time over which AI models can work is steadily increasing as models get better: this measure has now become a key indicator of AI progress. We’re able to complement METR’s analysis using our economic primitives. In the graph below, we show Claude’s task-level success rates relative to the amount of time a human would take to do the same task, both on Claude.ai and on our API: Task success vs. human-only time. This chart shows the relationship between task success (%) and the time the task would require a human to complete alone, all measured at the O*NET task level and split by platform. The dashed lines show the fit from a linear regression. METR’s benchmark suggests that Claude Sonnet 4.5 (the model in our own analysis) achieves 50% success rates on tasks of 2 hours. By contrast, our own API data finds that Claude is 50% successful at tasks that take nearly twice as long (around 3.5 hours), and on Claude.ai, the duration is vastly longer still—around 19 hours. But this might not be as discordant as it seems: our methodology is different to METR’s in some important ways. In our sample, users can break down complex tasks into smaller steps, creating a feedback loop that allows Claude to correct course. And rather than a fixed set of tasks, our sample contains a form of selection bias: users bring tasks to Claude that they’re more confident will work. Our analysis shows how Claude’s effective time horizons might look different to those found in a study with a consistent set of tasks. We’ll track this indicator in further reports. How does the nature of Claude’s work vary across countries? We find that Claude completes very different kinds of tasks in countries at different stages of economic development. In countries with higher GDP per capita, Claude is used much more frequently for work or for personal use—whereas countries at the other end of the spectrum are more likely to use it for educational coursework. This fits a straightforward “adoption curve” story, in which lower-income countries show a large share of AI use on education and on a smaller number of work tasks, while AI use diversifies towards personal purposes as countries become richer. These results align with recent work by Microsoft that associates AI use in education with lower per-capita income, and AI use for leisure with higher incomes. Our recent partnership with the Rwandan government and ALX, a technology training provider, is designed with this in mind: participants begin by developing AI literacy, and we’re piloting a…