Could AI lift US GDP to $44.4 trillion by 2030?

Anthropic's $44.4 trillion US GDP scenario requires recursive AI improvement, fast adoption and automation of most knowledge work. It is not a certain forecast.

Main report

On October 17, 2026, Anthropic co-founder Jack Clark spoke at the IMF–World Bank annual meetings in Bangkok about the pace of AI progress and its possible economic effects. Reuters reported that an internal Anthropic model produces a $44.4 trillion US GDP outcome for 2030 under a particularly strong set of assumptions. The previous day, the Anthropic Institute published “The AI expectations gap” and an interactive economic scenario model.

$44.4 trillion is a scenario, not a guaranteed forecast

Anthropic presents three paths. The modest scenario puts 2030 US GDP at $34.1 trillion, the substantial scenario at $36.3 trillion and the extreme scenario at $44.4 trillion, all in 2025 prices. Relative to the model's baseline, the AI-related differences are 1.6%, 8.3% and 32.4%.

Reuters' “almost 50%” comparison contrasts $44 trillion with the current economy of just over $30 trillion. It does not mean Anthropic has demonstrated that AI will certainly add 50% to GDP. Ordinary growth through 2030 and the model's assumptions are also part of the final level.

Analysis

What the extreme scenario assumes

The $44.4 trillion result requires AI to outperform people on the vast majority of knowledge-work tasks, perform nearly all of them autonomously, diffuse rapidly through business and create almost no new knowledge tasks for people. Anthropic says this path would likely require recursively self-improving AI and could push annual GDP growth toward 15%.

A wide expectations gap

The study surveyed 10,980 US adults and Anthropic employees. Ninety-one percent of Anthropic staff expect AI to be capable of a Nobel-level discovery by 2030, compared with 39% of the public. Staff estimate on average that AI could perform 75% of their current work by 2030, versus 28% among US workers.

The typical public respondent implies a much less extreme outcome: 2030 GDP about 10% above a no-AI counterfactual and unemployment around 5%. Anthropic itself says the model is not a complete map of reality.

What to watch

How businesses should use the research

The useful contribution is the set of variables to measure: the share of tasks AI can perform, actual business adoption, autonomy, productivity and the time workers need to move into new roles. A company should not budget around the $44.4 trillion scenario. It should measure time saved, usable output, review cost and error rates for each workflow.

Survey respondents identified customer preference for human interaction, privacy and data security, AI reliability, laws and liability among the main adoption barriers. These can make real deployment slower than technical capability.

Read the pillar analysis on AI infrastructure's trillion-dollar returns test and the related article on AI, growth and jobs. This is scenario analysis, not a certain forecast or investment advice.

Verified sources: the Anthropic Institute study, Anthropic's economic scenario explorer, the official IMF event page and Reuters, October 17, 2026.