Anthropic's Economics team has built a public tool that lets Americans explore three starkly different economic futures shaped by artificial intelligence, revealing that the impact of AI on jobs, growth, and unemployment depends entirely on how capable AI becomes and how quickly companies adopt it. The research, published today, challenges both utopian and dystopian narratives by showing that outcomes fall on a spectrum, and that society's policy choices will determine whether AI's gains are broadly shared or concentrated.

What Are the Three Economic Scenarios Anthropic Modeled?

Anthropic's team created three distinct pathways for how AI might reshape the U.S. economy by 2030. Each scenario reflects different assumptions about AI's capabilities, adoption rates, and how much work AI will do autonomously versus augmenting human workers.

• Modest Scenario: AI has roughly the same economic impact as the internet did, driving real gains within historical norms for new technologies and arriving gradually over time.
• Substantial Scenario: AI becomes capable of doing half of all knowledge work by 2030, mostly autonomously, but adoption remains incomplete. The economy grows at twice its normal rate, wages for knowledge workers stay flat, and other workers see gains.
• Extreme Scenario: AI becomes more productive than humans at the vast majority of knowledge-work tasks, does nearly all of them autonomously, and creates essentially no new knowledge tasks for people. Annual GDP growth reaches 15% per year, doubling the economy every 4.5 years, but unemployment rises beyond typical recessionary levels and far fewer workers have jobs in knowledge work.

The research is grounded in a technical report titled "Economic Scenarios for Transformative AI," authored by Korinek and colleagues in 2026. Rather than making a single prediction, Anthropic built an interactive explorer that lets the public input their own assumptions about AI capabilities and adoption, then see what economy those beliefs would produce.

How Does AI Transform Individual Jobs and Tasks?

The model underlying these scenarios doesn't treat jobs as monolithic units. Instead, it breaks every occupation into discrete tasks. A nurse, for example, performs rounds to check on patients, draws blood, triages incoming patients, charts vitals, and orders supplies. AI can affect each task differently.

Anthropic's framework identifies four ways AI interacts with work. First, AI can leave tasks unchanged. Second, it can augment workers by helping them do tasks better or faster, such as drafting discharge instructions or planning a shift's care schedule. Third, AI can automate tasks entirely, like charting vitals or ordering supplies. Fourth, AI can create entirely new tasks that didn't exist before, such as reviewing an AI-proposed care plan or monitoring how well an AI triages patients.

This task-based approach matters because it shows how productivity gains emerge. As a nurse incorporates AI into her workflow, she can oversee and accomplish more work. She spends more time talking with patients and helping them understand diagnoses. Productivity increases, but the job itself evolves rather than disappearing.

What Do Americans Actually Expect to Happen?

Anthropic surveyed more than 10,000 Americans in August 2026 about their views on present and future AI capabilities, adoption rates, and job transition difficulty. The results reveal a public that leans toward the "substantial change" scenario, not the extreme one.

The typical respondent's answers imply that GDP will be 10% higher by 2030 than it would be without AI, and the overall unemployment rate will rise to around 5%. However, around 10% of respondents hold views aligned with the extreme scenario, suggesting a meaningful minority expects transformative disruption.

The survey asked five specific questions to build these predictions. Respondents estimated how many tasks AI could perform, how much people would actually use AI, how much AI would work autonomously versus with human oversight, how much more productive AI would make workers, and how long it would take displaced workers to find new jobs. The variation in answers across the 10,980 general public respondents and 5,076 site visitors shows that Americans hold genuinely different beliefs about AI's trajectory.

Steps to Understand Your Own AI Economic Predictions

Anthropic designed the scenario explorer to be interactive and accessible to non-experts. Here's how to use it to test your own assumptions about AI's economic future:

• Input Your Capability Estimate: Start by deciding what percentage of tasks you think AI will be able to perform by 2030. This ranges from 0% to 100% and anchors all downstream predictions.
• Estimate Adoption Rates: Consider how widely companies and workers will actually deploy AI for the tasks it can do. High capability doesn't guarantee high adoption; regulatory, cultural, or economic barriers might slow rollout.
• Assess Autonomy Levels: Decide whether AI will work mostly under human supervision or operate independently. This distinction dramatically affects how many jobs remain and how much human oversight is needed.
• Project Productivity Gains: Estimate how much more productive AI will make individual workers. Options range from no change to a 10-fold increase in output per worker.
• Evaluate Job Transition Speed: Estimate how quickly displaced workers can find new employment. This ranges from one to two months to three or more years, and affects how painful the transition is for affected workers.

Once you input these five variables, the explorer calculates what the 2030 economy would look like under your predictions and compares your views to those of other respondents.

Why Does Anthropic Care About Economic Scenarios?

The company frames this research as part of a broader commitment to ensuring that AI's economic transition benefits society, including workers. Anthropic noted that while its Economic Index measures how AI is being used across the economy right now, the scenario explorer focuses on looking ahead and building visibility into possible futures.

The stakes are high. In scenarios where growth remains within historical norms, unemployment stays within the historical range and wages remain flat or rise depending on the industry. But in scenarios where growth accelerates to unprecedented levels, driven by recursively self-improving AI systems and rapid adoption, the distributional consequences become severe. Society becomes far wealthier overall, but the challenge shifts to ensuring those gains are broadly shared rather than concentrated among AI owners and highly skilled workers.

Anthropic emphasizes that the future depends on both technical progress and human choices. How AI's capabilities advance, how industries adopt those capabilities, and what policies governments put in place will all shape which scenario actually unfolds. By providing better visibility into possible economic futures, the company argues, society can take deliberate steps to steer toward more favorable outcomes.

The research arrives at a moment when AI's economic impact remains uncertain. Over $30 trillion of value is created annually in the U.S. economy through the combined work of people and the machines and software they use. How AI reshapes that value creation, and whether the benefits flow broadly or narrowly, will define the next decade of economic policy and social stability.