Anthropic has released an interactive economic modeling tool that allows users to test how different artificial intelligence assumptions could affect the US economy. The scenarios range from a modest productivity increase to GDP growth exceeding seven times the current pace alongside unemployment of 14%. Users can adjust AI capability, business adoption speed, whether AI supports or replaces workers, and how quickly displaced employees find new jobs. The model incorporates data from the Anthropic Economic Index (AEI), a research program analyzing millions of anonymized conversations to track real-world AI use across industries and geographies. AEI data found directive and automated AI use increased from 27% to 39% in early measurement periods, led by business and API users. Average task completion was about 80% faster, while complex tasks were completed 9 to 12 times faster. The model estimates that AI could raise US labor productivity growth by roughly 1% to 1.8% annually, compared with an average of about 1.4% over the past two decades. Anthropic co-founder Jack Clark has identified the pace of AI diffusion and the accompanying policy response as critical factors. A survey of nearly 11,000 people found broad agreement that AI could lift productivity but sharply divided views on its net effect on employment. AEI research also found that regions with higher GDP per capita tend to have higher automation adoption, indicating that AI’s economic effects may be unevenly distributed.