Last Updated: 11 September 2026
Anthropic's economics team published Economic Scenarios for Transformative AI on 9 September 2026, and it puts hard numbers on a question every business owner is asking. In the three scenarios modeled, US GDP in 2030 lands between 1.6% and 32.4% above a comparable economy without AI, unemployment ranges from historical norms to nearly 12%, and labor's share of national income falls from about 60% to as low as 45%. The finding that matters most for trades, construction and allied health: wages for physical occupations rise in every scenario, while the disruption concentrates in screen-based knowledge work.
This article breaks down what the report actually says, what independent data already shows, and what trade and construction businesses should do about it before 2030.
What Is Anthropic's 2030 Economy Report?
Economic Scenarios for Transformative AI is a technical report by Anthropic's Economics Team (Korinek et al., 2026) released alongside an interactive Econ Scenario Explorer on 9 September 2026. It models the US economy from 2026 to 2030 across three scenarios, mapping paths for AI capability and adoption onto GDP, unemployment, wages, job reallocation and labor's share of income. According to Anthropic, it is explicitly a scenario planning tool, not a forecast.
The model treats occupations as bundles of tasks. Some tasks get augmented by AI, some get automated, and new tasks emerge. Five inputs drive everything: how capable AI becomes, how widely it is adopted, how autonomously it works, how productive it makes people, and how quickly displaced workers find new occupations. Anthropic's economist Peter McCrory told Axios: "Part of the value of doing scenario modeling is so that you can do scenario planning."
"The future is not predetermined." Anthropic, Scenarios for our Economic Future, September 2026
What Are the Three Scenarios for the 2030 Economy?
Anthropic models a modest scenario where AI's impact resembles the internet's, a substantial scenario where AI handles roughly half of knowledge work autonomously, and an extreme scenario where recursively self-improving AI takes over most cognitive work. 2030 GDP comes in at $34.1 trillion, $36.3 trillion or $44.4 trillion in 2025 dollars respectively, which is 1.6%, 8.3% or 32.4% above a no-AI baseline.
| Metric (2030, US) | Modest | Substantial | Extreme |
|---|---|---|---|
| GDP (2025 dollars) | $34.1T | $36.3T | $44.4T |
| GDP vs no-AI baseline | +1.6% | +8.3% | +32.4% |
| Overall unemployment | Historical range | 4.6% | ~12% |
| Knowledge worker wages | Rise modestly | Roughly flat | Down 10%+ |
| Labor share of income | 59.4% | 56.1% | 45.2% |
According to Anthropic's August 2026 survey of 10,980 Americans, the typical respondent's expectations about AI capability and adoption imply an outcome close to the substantial scenario: GDP about 10% higher by 2030 than it would be without AI, with unemployment around 5%. Around 10% of respondents hold views consistent with the extreme scenario.
Will AI Cause Mass Unemployment by 2030?
Only in the extreme scenario. According to Axios reporting on the model, overall unemployment reaches 4.6% in the substantial scenario and nearly 12% in the extreme case, with unemployment among knowledge workers hitting 17.9%. In the modest and substantial scenarios, unemployment stays within ranges the US economy has seen before. The dominant labor market effect in most scenarios is occupational churn: workers changing fields rather than permanent joblessness.
The mechanism matters for planning. Coders and call centre workers in the model move toward occupations like electrician and nurse, which are less exposed to AI. But switching occupations is slow and hard, so the more switching a scenario requires, the more people sit between jobs. The unemployment risk is a transition cost, not a permanent state, and policy responses (which the model excludes) would shape it.
Which Jobs Are Safest from AI by 2030?
According to Anthropic's model, the safest occupations are physical and manual ones: construction trades, electrical work, nursing and other hands-on roles. Wages outside knowledge work rise in all three scenarios because AI productivity in cognitive tasks increases demand for the physical work it supports. Faster design and permitting, for example, means more construction projects get built, which pushes construction wages up.
US Bureau of Labor Statistics projections through 2034 point the same direction. According to the BLS, data scientist employment is projected to grow 33.5% and software developer employment 15.8%, while customer service representative employment falls 5.5% and administrative assistant employment falls 5.8%. Technical and physical roles grow while routine information-processing roles shrink.
Why Does Labor's Share of Income Fall?
Today about 60 cents of every dollar the US economy produces goes to workers and 40 cents to capital. In Anthropic's scenarios, labor's share falls to 56.1% (substantial) or 45.2% (extreme) because AI makes capital more useful for more tasks, raising demand for it relative to human labor. Average wages can rise across the economy at the same time, because the pie grows faster than labor's slice shrinks.
The strategic implication for established businesses is blunt: the share shift rewards owners of capital, and an AI-augmented business is exactly that. Business owners who adopt AI early capture productivity gains on the capital side of the ledger. Workers and firms that sell raw labor hours face the squeeze. This is the quiet story in the report, and it is the one worth acting on.
What Does Current Jobs Data Say About AI Right Now?
According to Stanford's Digital Economy Lab, using ADP payroll data through June 2026, there is no evidence yet of economy-wide AI job displacement. But employment for workers aged 22 to 25 in highly AI-exposed occupations sits about 19% below where it would be if it had kept pace with less-exposed peers. The gap comes from reduced hiring of young workers, not higher firing rates.
That is the leading edge of the transition: organisations quietly stop hiring for exposed tasks and let attrition do the rest. If you are waiting for a headline event before acting on AI, the Stanford data says the shift is already underway, silently, in hiring pipelines.
How Does Anthropic's Forecast Compare with McKinsey and Goldman Sachs?
Anthropic's extreme scenario implies roughly $10.8 trillion in added US GDP by 2030. For comparison, McKinsey has projected AI could add about $13 trillion to the global economy by 2030, and Goldman Sachs research puts the boost from widespread adoption at roughly $7 trillion in annual global GDP over a decade, with some of its economists modeling closer to $1 trillion per year. No other major forecast pairs a US-only number that large with a double-digit unemployment rate, which is what makes Anthropic's scenario range notable.
What Should Trade and Construction Businesses Do Before 2030?
The play for trades and construction is the inverse of the doom narrative. Your trade gets more valuable in every scenario, and the knowledge work wrapped around it, quoting, scheduling, job documentation, compliance and invoicing, gets cheaper to automate. The winners compress the second category and reinvest the time in capacity and customers. Here is a four-step version.
- Audit where knowledge work sits in your business. List every task that happens on a screen, score each by hours per week and cost per hour, and rank them. That list is both your exposure map and your automation pipeline.
- Automate quoting and scheduling first. They are the highest-volume, highest-friction knowledge tasks in most trade businesses, and faster quote turnaround directly lifts enquiry-to-job conversion.
- Point freed admin time at revenue, not headcount cuts. Demand for the physical side of your business rises in these scenarios. Capacity, follow-ups and customer experience compound while competitors automate nothing.
- Track adoption metrics monthly. Quote turnaround time, admin hours per job, jobs per week and gross margin per job tell you whether the automation is actually converting into the productivity the scenarios describe.
The Flowtivity Take: The Bottleneck Is Adoption, Not Capability
The gap between Anthropic's modest and extreme scenarios is not really about AI capability. It is about adoption speed and how quickly workers and businesses adjust, which are organisational problems, not technical ones. That gap is where a small business either captures the productivity gains or watches them accrue to whoever adopts faster.
In Flowtivity's AI opportunity work with Australian trade, construction and allied health businesses, the pattern is consistent: the highest-return automations are almost never the trade work itself. They sit in quoting, scheduling, job documentation, compliance and invoicing, exactly the task categories Anthropic's model flags as automatable. We have delivered AI training for Occupational Therapy Australia, and across AJ Awan's consulting career, including 6 years at EY as Manager in IT Advisory, engagements have delivered an average of $15 million in business benefits per engagement.
Anthropic closes its own report with the line that matters: the future is not predetermined. For trade and construction owners, that cuts both ways. The scenarios say your trade skills appreciate. Whether your business captures the knowledge-work savings around them is a 2026 decision, not a 2030 one.
About the author: AJ Awan is the founder of Flowtivity. Former EY management consultant. TOGAF certified enterprise architect. Flowtivity helps established Australian businesses in trades, construction, allied health and professional services find and automate their highest-ROI workflows with AI.
