Anthropic Releases Interactive Model of AI’s Possible Economic Futures

Anthropic’s Economics team released a new interactive model in September 2026: the Econ Scenario Explorer, which projects how AI could affect US growth, jobs, wages, and unemployment through 2030, alongside a companion working paper and a survey of 10,980 US adults.

The explorer page identifies the release as version 1.0, dated September 2026. It is built on the technical report Economic Scenarios for Transformative AI, published as The Anthropic Institute Working Paper No. 2026-02. Anton Korinek, Charles I. Jones, Szymon Sacher, Tess Cotter, and Peter McCrory developed the model and co-authored the report; McCrory, Korinek, and Charles Yang coordinated the project, and Jack Clark provided direction. The authors note that they used Claude as a research and writing assistant. The Anthropic Institute presents the project as a new model of how AI may affect economic growth, jobs, wages, and more by 2030.

The release sits alongside Anthropic’s existing Economic Index. Where the index measures how AI is being used across the economy, the explorer looks ahead: users enter expectations for how capable AI will become and how widely it will be used, and the model shows the 2030 economy those expectations would imply, along with how a user’s answers compare with those of more than 10,000 surveyed Americans.

A Task-Based Model of the US Economy

The model represents every job in the economy as a bundle of tasks drawn from the US Department of Labor’s O*NET taxonomy. For each task, AI can augment a worker’s performance, automate the task outright, leave it unchanged, or create new tasks. Workers are employed either in cognitive occupations, meaning management, professional, sales, and office jobs whose tasks AI can affect, or in all other occupations; the explorer labels the two groups knowledge workers and all other workers. The cognitive group employed 62.4 percent of workers in the 2025 Current Population Survey.

According to the paper, a scenario is a path for five objects: the share of tasks AI affects, the diffusion of AI use across those tasks, the productivity gain per AI-performed task, the split between automation and augmentation, and the rate at which new labor tasks are created. When displaced cognitive workers must search for jobs in other occupations, labor-market frictions (a discount on cross-occupation search, the speed at which expanding occupations post openings, and rigidity of the cognitive wage) can produce sustained unemployment. The authors state that the scenarios are not predictions and that they attach no probabilities to them; the framework is intended as a structured way to compare possibilities under different assumptions.

Three Scenarios for 2030

In the modest scenario, the paper reports, GDP in 2030 is 1.6 percent above its no-AI path, growth runs at 2.4 percent a year against 2 percent without AI, unemployment stands at 3.9 percent against a normal level of 3.8 percent, and the labor share slips from 60.0 to 59.4 percent of income. The explorer expresses 2030 GDP at 2025 price levels as $34.1 trillion in this scenario.

In the substantial scenario, the paper reports, 2030 GDP is 8.3 percent above the no-AI path, or $36.3 trillion at 2025 prices, and growth over the twelve months to 2030 reaches 5.4 percent a year, a pace the paper sets against the 4.7 percent fastest GDP growth of the 1990s dot-com boom, recorded in 1999. Unemployment among cognitive workers rises from 2.9 percent in mid-2026 to 4.5 percent in 2030. Cognitive wages end the period 0.3 percent below their no-AI path while wages in all other occupations run 5.9 percent higher, and the labor share falls from 60 to 56.1 percent of income.

In the extreme scenario, 2030 GDP is 32.4 percent above the no-AI path, or $44.4 trillion, with growth of 15.4 percent a year; the explorer describes the economy doubling in size every 4.5 years on this path and says it would likely require recursively self-improving AI systems and rapid adoption for knowledge work. Cognitive-worker unemployment reaches 17.9 percent and economy-wide unemployment 11.9 percent. Cognitive wages fall 11.5 percent below the no-AI path while other wages rise 33.6 percent above it, and the labor share drops from 60 to 45.2 percent of income as capital’s share rises to 54.8 percent.

The paper states that in the extreme scenario a transfer of about 9 percent of GDP, roughly the size of Social Security and Medicare combined, would hold cognitive workers’ income at its no-AI level, and it notes that there is no precedent for transfers of that scale in response to technological change.

What Americans Expect

Anthropic surveyed a representative sample of 10,980 US adults through Morning Consult, fielded August 11–23, 2026. Respondents answered five questions: when AI will perform each of eight tasks as well as a skilled professional, how widely it will be used, how large its productivity gains will be, whether it will automate or augment work, and how long a displaced worker would take to find a job in a new occupation.

The paper reports that running the median respondent’s answers through the model yields outcomes close to the substantial scenario: GDP 8.6 percent above the no-AI path by 2030 and unemployment around 4.6 percent. The explorer summarizes the typical respondent as implying GDP about 10 percent higher by 2030 than without AI and overall unemployment around 5 percent, with roughly 10 percent of respondents holding views in line with the extreme scenario. In the survey results, 53 percent of respondents said AI can already write routine business emails and documents, and 24 percent said it can already build and maintain a working software product; 39 percent expect a Nobel-level scientific discovery by 2030, while 40 percent say that will never happen.

Stated Limits and External Review

Anthropic describes the explorer as Version 1.0 and cautions that it isolates a few key forces while omitting many others: it leaves out policy responses, business cycles, potential aggregate-demand or financial-market disruptions, and possible catastrophic risks, and it includes no scenarios with hyper-capable robots. The paper states that the framework considers only AI’s effects on cognitive tasks and does not allow for rapid advances in robotics affecting physical tasks, largely the reason the analysis ends at 2030.

Outside economists, including Daron Acemoglu, David Autor, and David Romer, commented on an early draft of the report; Anthropic says they were not asked to endorse its conclusions. The explorer page lists criticisms that remain open, among them that the model does not follow individual workers and therefore captures the costs of job displacement only coarsely, and that it omits the aggregate-demand effects of the data-center buildout. Anthropic says it plans to address many of the criticisms in future versions.

Anthropic says the model, alongside its full research portfolio, will inform the research it funds through its Economic Futures program on interventions for labor-market disruption, as well as the policy ideas it proposes. The authors write that data observed over the coming years will indicate which scenario the economy is in, and that none of the three scenarios can be ruled out.