Anthropic's latest economic scenarios suggest that artificial intelligence could reshape the U.S. economy far beyond simple automation. In the company's most ambitious model, AI lifts productivity so strongly that national output could rise by about 33% by 2030, even as many knowledge-based roles lose value.
The research, published by the Anthropic Institute, explores three possible paths for the U.S. economy. In the mildest version, AI behaves like a useful tool and adds only a small boost to GDP. In the middle scenario, AI is used across a meaningful share of tasks, pushing growth higher while reducing demand for some office and professional jobs. In the most aggressive case, AI handles a large portion of knowledge work and delivers a dramatic jump in productivity.
Anthropic frames the findings as scenarios, not predictions. The key variable is whether AI mainly assists workers or increasingly replaces them. The model treats jobs as bundles of tasks, meaning some duties can be automated while others remain distinctly human.
That distinction matters across sectors. Physical and on-site work such as construction, electrical work, and similar trades could become more valuable, while programmers, accountants, lawyers, analysts, and managers may face pressure on wages and employment. In the strongest scenario, the model shows labor's share of income falling while capital captures a larger portion of the gains.
The broader message is not that AI will simply erase work, but that it may redistribute opportunity in uneven ways. The future labor market could reward skills that are harder to automate, while pushing organizations to rethink training, entry-level roles, and career pathways.
If these trends continue, AI may become one of the defining engines of economic change, shaping not only how much societies produce, but also how value is shared in the decades ahead.
