Anthropic Model Projects AI Could Boost U.S. GDP by Over 30% or Drive Historic Unemployment by 2030

Anthropic, the developer of the Claude AI assistant, unveiled an interactive economic model designed to project how artificial intelligence could reshape the U.S. economy by the end of the decade. The tool, built by the company’s economics team, breaks jobs into bundles of tasks and estimates the impact of automation, augmentation, and newly created work across different levels of AI capability and adoption.

The model runs three core scenarios — modest, substantial, and extreme — each projecting a different trajectory for gross domestic product, wages, and employment.

Under the modest scenario, where AI provides a marginal boost and the economy continues on a near-normal path, U.S. GDP in 2030 reaches $34.1 trillion, a 1.6% increase over baseline, according to Anthropic’s technical report. Unemployment stays within the historical range, and wages remain flat or rise depending on the industry.

In the substantial scenario — which corresponds to the average expectations of roughly 11,000 Americans surveyed by Anthropic — GDP rises 8.3% to $36.3 trillion by 2030. The overall unemployment rate ticks up to around 5%, and some knowledge workers face displacement. The survey, conducted in August, found that the typical respondent’s answers implied an outcome close to this scenario. About 10% of respondents held views consistent with the extreme scenario.

The extreme scenario projects the most disruptive outcome: GDP climbs 32.4% to $44.4 trillion, an economy Anthropic says would be 50% larger than today, as reported by Gizmodo. But growth comes with severe labor-market pain. Gizmodo reports the model shows unemployment reaching nearly 30%. NPR, citing the same model, puts job losses at nearly 14% of workers, with fewer than half finding new employment. In both readings, the disruption falls heavily on knowledge workers as AI automates tasks previously done by coders, call-center agents, and similar roles.

“In those scenarios, society is far wealthier, so the challenge is making sure that the gains are broadly shared,” Anthropic wrote in a blog post.

The model does not account for a potential AI investment bubble or its aftermath, per Gizmodo. Nor does it factor in the risk that AI could cause human extinction — a possibility that Evan Hubinger, Anthropic’s alignment science lead, publicly put at greater than 10% over the next decade, according to Gizmodo. Hubinger stated, “We really do earnestly believe AI could kill all humans,” adding that the company does not yet have a plan to solve alignment for superintelligence. The economic model does not include that outcome.

Anton Korinek, Anthropic’s head of transformative AI economic studies, told NPR that a key variable is adoption speed. “If the AI can do amazing things but nobody uses it, then it’s not going to have an economic impact,” he said.

Anthropic co-founder Jack Clark told NPR he expects technology to improve rapidly but diffusion through the economy to be slow. “It will get really, really good. But it will make its way into the economy more slowly,” Clark said. He added that extremely high GDP growth could give policymakers fiscal breathing room: “If you end up with this level of GDP growth, you have moves available to you as a [government] policymaker that are unimaginable today. Policymakers should get ready to spend.”

The interactive tool allows users to input their own assumptions about AI capabilities, adoption rates, autonomy, productivity gains, and how quickly displaced workers find new jobs. The model then generates a custom 2030 projection for GDP, job reallocation, and wage shares.

The company positions the model as a resource for policymakers and academics, not as a single forecast. “We don’t know yet how AI will reshape the economy,” Anthropic stated. The authors declined to offer a prediction of which scenario is more likely, noting that AI experts tend to project faster adoption while economists are more cautious.