How to Understand AI’s Economic Future with Anthropic’s Scenarios

In September 2026, Anthropic’s Economics team released Scenarios for Our Economic Future, an interactive model of how AI might shape jobs, growth, and unemployment in the US through 2030.

It is not a normal forecast with one answer. Instead, it lets you:

• Read three distinct scenarios for the US economy
• See how each scenario shifts GDP, jobs, wages, and who gets the gains
• Enter your own predictions about where AI is heading
• See the 2030 economy that your predictions imply
• Compare your view with how other people answered

It builds on the technical report Economic Scenarios for Transformative AI (Korinek et al., 2026). Source: anthropic.com/institute/econ-scenarios

Use it to think through possible futures. It does not predict the future.

Anton Korinek, Chad Jones, Szymon Sacher, Tess Cotter, and Peter McCrory built the economic model and co-wrote the technical report.

Eighteen outside economists, among them Daron Acemoglu, David Autor, Pete Klenow, Ben Moll, David Romer, and Jón Steinsson, commented in detail on an early draft. Anthropic notes that it did not ask reviewers to endorse its conclusions.

The model treats every job in the economy as a bundle of tasks, drawn from the US Department of Labor’s O*NET task lists. AI can affect each task in one of four ways:

Unchanged

AI leaves the task as it is.

Augmented

AI helps a person do the task faster or better.

Automated

AI performs the task on its own.

New

AI creates tasks that never existed before.

Jobs rarely just vanish or survive. Their task bundles change.

Anthropic uses a nurse to show one job splitting across all four outcomes:

Only humans

Bathing a patient

Augmented

Drafting discharge instructions, monitoring patients remotely, and planning the shift’s care schedule

May be automated

Charting a patient’s vitals and ordering the ward’s supplies

New tasks

Checking how well an AI triages patients and reviewing an AI-proposed care plan

The outcome: the nurse oversees more, gets more done, and has more time to talk with patients. Productivity goes up.

Add up every task done in the US, by people and by the machines and software they use, and you get the US economy: more than $30 trillion of value created in the past year.

How AI reshapes that economy depends on:

• Which tasks AI is able to do
• Whether AI augments or automates those tasks
• How much more productive AI makes people
• How fast workers and companies adopt AI
• Which new tasks AI creates

Modest

An impact like the internet’s. Hard to spot in macroeconomic data.

Substantial

A larger impact than the internet or the railroad.

Extreme

A fully transformed, unprecedented economy, likely driven by recursively self-improving AI and faster adoption.

AI has about the same kind of impact the internet had. The gains are real, but they arrive gradually and stay within the historical norm for new technologies.

GDP in 2030

$34.1 trillion, 1.6% higher

Labor share

59.4% to workers and 40.6% to capital

By 2030, AI can do half of all knowledge work, most of it autonomously. Adoption lags capability, so most knowledge work tasks are still done without AI.

GDP in 2030

$36.3 trillion, 8.3% higher

Growth

The economy grows at double its normal rate

Wages

Knowledge workers’ wages stay about flat. Other workers gain.

Labor share

56.1% to workers and 43.9% to capital

AI outperforms humans at the vast majority of knowledge-work tasks, handles nearly all of them autonomously, and creates almost no new knowledge tasks for people.

GDP in 2030

$44.4 trillion, 32.4% higher

Growth

Annual GDP growth hits 15%, so the economy doubles in size every 4.5 years

Jobs

Far fewer people work in knowledge work, and unemployment climbs past typical recessionary levels

Wages

Knowledge workers’ wages drop by more than 10% by 2030

Labor share

45.2% to workers and 54.8% to capital

Anthropic says this scenario would likely need recursively self-improving AI, adopted fast for knowledge work.

GDP rises in all three scenarios, but by very different amounts: from 1.6% to 32.4% higher by 2030. Every GDP figure uses 2025 price levels.

Growth is not in doubt. How it is shared is.

The job market always has churn. In the substantial and extreme scenarios, knowledge workers face far more automation and displacement.

Anthropic’s example: coders and call center agents may need to move into jobs like electrician and nurse, which are less exposed to AI.

Changing occupations is hard because people:

• May not want to leave their occupation
• May need to learn new skills first
• Still need time to land the next job

In most scenarios, job reallocation and unemployment stay within historical ranges. In the extreme scenario, unemployment could spike to historic highs.

Average wages go up in all three scenarios, but most of the rise lands outside knowledge work.

• Lower demand for human knowledge work pushes those wages down
• Workers need time to move into occupations with rising demand
• Faster designs and permits can lead to more construction projects
• Higher demand for construction workers pushes their wages up

Knowledge workers’ wages stay about flat in the substantial scenario and drop by more than 10% in the extreme one.

Today, roughly 60 cents of every dollar the economy produces goes to workers and roughly 40 cents goes to capital.

Modest

Capital share rises 0.6 points

Substantial

Capital share rises 3.9 points

Extreme

Capital share rises 14.8 points

In the extreme scenario, total labor income barely moves by 2030, even though the economy is much larger.

Anthropic: the hard part is not reaching growth but spreading its benefits widely.

In August 2026, Anthropic and Morning Consult surveyed 10,980 Americans about AI’s capabilities, its adoption, and how easy it would be to find new work.

• The typical answer points to outcomes near the substantial scenario
• That implies GDP about 10% higher by 2030 than without AI
• It also implies overall unemployment of around 5%
• About 10% of respondents hold views that match the extreme scenario

Capabilities

What share of tasks is AI able to do?

Adoption

How much do people actually use AI?

Autonomy

How much does AI do on its own, from almost none to almost all?

Productivity

How much more productive does AI make people, from no change to 10 times or more?

Adjustment

How long do people take to find a new job, from one to two months to three years or more?

Your answers generate your own version of the 2030 economy.

Anthropic calls the model a stark simplification. Version 1.0 does not include:

• Hyper-capable robots
• Policy responses
• Business cycles
• Aggregate demand and financial market disruptions
• Catastrophic risks
• Demand effects from the data center buildout
• Individual workers’ paths, so it captures displacement costs only roughly

Reviewers also disagreed at the margins. Some saw the extreme scenario as a thought experiment, some felt the modest one understates what is already visible, and some argued AI could speed up technological progress more than the model assumes.

Real outcomes may differ materially. Use the scenarios to think, not to bet.

Apply the model to your own work. List what you really do in a normal week:

• Recurring tasks
• Writing and documents
• Analysis and decisions
• Meetings and conversations
• Hands-on or physical work
• Relationships with clients or colleagues
• Tasks you supervise or review

Be concrete. “Write the weekly sales report” is a task. “Marketing” is not.

For each task, ask whether AI will most likely leave it unchanged, augment it, automate it, or create new work around it.

Unchanged

Protect these strengths and deepen them

Augmented

Learn to do these faster and better with AI

Automated

Plan for these to shrink

New

Move toward these early

Open the scenario explorer and answer the five questions honestly. Next:

• Note which scenario your answers sit closest to
• Compare your view with the survey results
• Change one answer at a time to see what matters most
• Write down the reason behind each answer

Revisit your answers every few months as new evidence comes in.

In every scenario, people who use AI well on augmented tasks are in a stronger position. Practice:

• Giving AI clear context and clear goals
• Reviewing and fixing AI output
• Checking facts and sources
• Supervising multi-step AI work
• Explaining AI-assisted results to other people

The model shows that the time it takes to find new work drives unemployment. Cut yours down:

• Keep an up-to-date record of your skills and results
• Learn one adjacent skill every year
• Build relationships beyond your current role
• Watch which nearby occupations are growing
• Keep a financial buffer where you can

Scenarios only help if you keep updating them. Watch:

• Anthropic’s Economic Index, which measures how AI is used across the economy today
• Job postings and wages in your field
• How fast your own employer adopts AI
• How much of your work AI handles by itself
• Official unemployment and wage data

Act as my career analyst. My job is [role] in [industry]. These are the tasks I do in a typical week: [list]. Classify each task as unchanged, augmented, automated, or new-task opportunity, based on what AI can do today. Explain your reasoning briefly and flag anything you are unsure about. Then suggest three skills to build and two adjacent roles to explore. Do not invent statistics.

• Treat any scenario as a forecast
• Quote the extreme scenario as a certainty
• Ignore the scenario that seems least likely to you
• Assume your job is safe because AI cannot do all of it
• Assume your job is gone because AI can do part of it
• Share the numbers without noting they come from a model