Artificial Intelligence can make the economy dramatically richer while simultaneously making cognitive skill workers substantially poorer by 2030, argues a new research study by Anthropic.
The paper titled “Economic Scenarios for Transformative AI” states that AI could have widely different effects on the economy by 2030, ranging from relatively limited productivity grains to major economic disruption.
The Anthropic paper presents three scenarios- modest, substantial and extreme- to assess how different levels of AI capability, adoption and automation could affect GDP, wages, employment and unemployment through 2030.
According to Anthropic’s researchers, under modest changes, AI adds less than half a point to GDP growth in US by 2030 with unemployment raised by tenth of a point.
In substantial scenario, AI affects roughly 12% of all economic tasks by 2030, with productivity on affected tasks rising about 57%, and 75% of AI-performed instances being automated rather than augmented. GDP ends up 8.3% above the no-AI path.
According to Anthropic’s study, the third and most extreme scenario, AI will have transformative effects with the technology performing half of today’s cognitive work by 2030.
“GDP growth then rises to 15 percent per year, the labor share of income falls from 60 to 45 percent, and nearly one in five cognitive workers is unemployed,” observes the report.
Researchers at Anthropic also remarked that AI would increase productivity while also displacing workers in cognitive occupations such as management, professional, sales and office work. The authors discovered that almost all of the divergence between the three scenarios occurs after 2027.
Researchers warned that despite probability of an extraordinary GDP boom of 15.4% by 2030 in ‘extreme scenario’, it does not mean everyone gets to win due to AI. The average wage is expected to be 9.7% above the no-AI path yet the wage for cognitive occupations is 11.5% below its no-AI path. Meanwhile, wages in other occupations are 33.6% higher.
The biggest distributional shift is from labor to capital. The labor share of income falls from 60% to about 45%, while the capital share rises from 40% to roughly 55%. In other words, a huge chunk of the additional economic value goes to owners of capital rather than workers.
The extreme labor-market numbers are brutal. Cognitive employment falls 21.5% from its mid-2026 level; 17.9% of workers who began in cognitive occupations are unemployed, while overall unemployment reaches 11.9%.
The report has also examined how Americans expect AI to develop. The researchers surveyed 10,980 US adults through Morning Consult between August 11 and 23, 2026. Participants were asked when AI will be able to do each of eight different cognitive tasks of increasing difficulty, how widely AI will be used, how large the productivity gains will be, whether workers will be automated or augmented, and how long a displaced worker will take to find a new job.
The median responses from the survey were found to be close to the substantial change scenario. Under that scenario, GDP is 8.3 percent above its no-AI path by 2030 and cognitive employment declines by 3.9 percent.
The authors also highlighted the potential policy implications if AI produces large productivity gains while significantly reducing demand for labour. In an extreme scenario they said around 18 percent of the cognitive labour force could be unemployed and a substantial share of income could shift from payrolls to capital.
However, the report has cautioned that its framework leaves out several factors such as catastrophic risks, political economy considerations, business cycles, and possible financial market disruptions. “Our framework omits many potentially relevant forces, for example, catastrophic risks, political economy considerations, business cycles, and possible financial market disruptions. All of these could be important in any of our scenarios,” the report mentioned.
Also Read: Anthropic Names Most AI-Exposed Jobs; Gap Between Capability and Reality Is Vast
