Anthropic Maps AI's Economic Upside – and Its Omissions

Anthropic, the developer of the Claude AI assistant, published a set of economic projections on Wednesday that map how artificial intelligence could reshape the U.S. economy by 2030 — from a modest productivity boost to growth "faster than anything in economic history." The company's economics team released a technical paper and an interactive tool that models the economy as bundles of tasks that AI can augment, automate, or create anew.

The authors stress that the scenarios are not forecasts. "The scenarios are not predictions and we attach no probabilities to them," the paper states.

**Three Scenarios, Three Outcomes**

In the **modest** scenario, AI functions as a minor technology. U.S. GDP in 2030 sits 1.6% above the no-AI baseline, annual growth reaches 2.4%, and cognitive employment — management, professional, sales and office work — falls by half a percent. Unemployment barely moves.

The **substantial** scenario doubles the economy's normal growth rate to 5.4% as AI becomes capable of half of all knowledge work, though most tasks are still done without its assistance. GDP lands 8.3% above baseline. Cognitive employment falls 3.9%, and unemployment among office workers rises to 4.5%. Wages diverge: cognitive pay dips slightly while other workers gain nearly 6%.

In the **extreme** scenario, annual growth hits 15.4%, GDP finishes 32.4% above the no-AI path, and the economy doubles roughly every four and a half years. But cognitive employment collapses by 21.5%, unemployment among those workers reaches 17.9%, and joblessness across the whole workforce hits 11.9% — worse than a typical recession. Office wages fall 11.5% while other wages jump 33.6%.

The starkest shift is who collects the proceeds. Labour's share of national income drops from 60% to 45.2%, with capital income rising more than 80%. "In those scenarios, society is far wealthier, so the challenge is making sure that the gains are broadly shared," Anthropic wrote.

**Public Expectations vs. CEO Warning**

Anthropic paired the model with a Morning Consult survey of U.S. adults fielded in August. According to the paper, the median respondent's expectations map onto the substantial scenario, implying GDP roughly 8% higher by 2030 and cognitive employment down about 4%. On the company's own site, Anthropic reported that the typical respondent among more than 10,000 Americans surveyed implied GDP 10% higher by 2030 and an overall unemployment rate around 5%.

Anthropic's own CEO sits squarely in the extreme scenario. In May 2025, Dario Amodei warned that up to half of entry-level office jobs could disappear within five years, with unemployment reaching 10% to 20%.

**Adoption, Not Just Capability**

The model treats adoption rates as a key variable. "If AI can do amazing things but nobody uses it, then it's not going to have an economic impact," said Anton Korinek, who leads Anthropic's transformative AI economic studies.

Anthropic co-founder Jack Clark told NPR he expects rapid technical progress but slower uptake. "I think the technology will keep developing at a very, very fast and sustained rate but diffusion of the technology will likely be more challenging than people think," Clark said. "So it will get really, really good. But it will make its way into the economy more slowly."

Clark added that high growth scenarios would give policymakers fiscal firepower: "If you end up with this level of GDP growth, you have moves available to you as a policymaker that are unimaginable today. Policymakers should get ready to spend."

**The Scenario That Is Not There**

Every path in the model assumes an economy that still functions. No scenario accounts for an AI bubble bursting, and none accounts for the catastrophic risks that some within the company have publicly raised.

The omission looked pointed this week as Jacob Coxon, a 27-year-old researcher who worked at both OpenAI and Anthropic, resigned on Tuesday and published a thread explaining why. "Neither company is acting responsibly," Coxon wrote, adding that "they are racing straight to self-improving superintelligence." He also claimed colleagues privately believe the technology "could kill us all by the end of the decade" while executives soften their language publicly, and described the industry's approach as "a hubristic gamble that should not be launched from a private company's Slack."

Separately, Evan Hubinger, Anthropic's alignment science lead, posted on Twitter that "We really do earnestly believe AI could kill all humans!" He put the odds at greater than 10% over the next decade while admitting "we do not yet have a plan to solve alignment for superintelligence and are not clearly on track to."

Anthropic has itself disclosed that Claude models gained unauthorised access to the real systems of three organisations this year. The economic model does not weigh these risks.