Anthropic Models AI Impact: 32.4% US GDP Growth Possible by 2030 in Extreme Scenario

ChainCatcher report: Anthropic’s economics team has built a model to analyze the impact of AI on U.S. employment, growth, and unemployment. The model treats the economy as a portfolio of tasks categorized by the U.S. Department of Labor’s O*NET system, where AI can preserve, augment, automate, or create new tasks. The U.S. economy is described as having over $30 trillion in task instance value. Three scenarios are defined based on AI capabilities and adoption speed: Under the moderate scenario—similar to the internet’s impact—U.S. GDP increases by 1.6% by 2030 (to $36.3 trillion in 2025 prices), with a labor share of 59.4% and a capital share of 40.6%. Under the substantial scenario, AI can perform half of all knowledge work, boosting GDP by 8.3% (to $36.3 trillion), with knowledge worker wages remaining largely flat and a labor share of 56.1%. Under the extreme scenario, AI becomes significantly more efficient than humans in nearly all knowledge work and completes it almost autonomously, driving GDP up by 32.4% (to $44.4 trillion), reducing knowledge worker wages by over 10%, lowering the labor share to 45.2%, and pushing unemployment above typical recession levels. A survey of over 10,000 Americans in August found that typical responses aligned closest with the substantial scenario (GDP about 10% higher in 2030, unemployment around 5%), with approximately 10% of respondents nearing the extreme scenario. In all scenarios, GDP rises; however, transformative scenarios require significant occupational transitions. The page provides access to the technical report and an interactive explorer.