Anthropic Co-Founder: Internal Model Shows AI Could Drive US GDP to $44 Trillion
Jack Clark, co-founder of Anthropic, stated that the company's internal economic model predicts that AI could significantly boost the US economic output to $44 trillion by 2030, up from just over $30 trillion currently.
Speaking at the IMF and World Bank annual meeting in Bangkok, he noted that this figure "sounds outrageous," but AI is greatly accelerating scientific discovery. Such outcomes could occur as long as technology continues to advance.
Clark added that if GDP growth is this substantial, economic theory suggests that entirely new industries and inventions need to be created to support such a scale of economic expansion. He also called for international cooperation to establish "extraordinary and serious rules" for this transformative technology.
Background information shows that in a previous economic scenario tool released by Anthropic, under extreme scenarios, US GDP could reach about $44.4 trillion, corresponding to approximately 32% growth, accompanied by a high unemployment rate among knowledge workers and an increase in capital share.
Clark leads the Anthropic Institute, which aims to publicly share internal laboratory hypotheses for external evaluation of the economic impacts of different AI capabilities and adoption paths.
Mechanically, the model assumes that rapid improvements in AI capabilities will permeate the economy, driving productivity and accelerating scientific discovery. Capital and technology providers may benefit first, while traditional knowledge jobs and labor-dependent industries face pressure, alongside a rising demand for regulation and international rule-making.
Additional observations indicate that the increase from about $30 trillion to $44 trillion corresponds to an average annual nominal growth of about 10%, emphasizing that the creation of new industries is a necessary condition, rather than merely efficiency improvements.
Source: Public information
ABAB AI Insight
As a co-founder of Anthropic, Jack Clark has long focused on AI safety and economic impact. In September 2026, the Anthropic Institute released an economic scenario tool that includes moderate, significant, and extreme paths, with the extreme scenario pushing US GDP to about $44.4 trillion by 2030, accompanied by a knowledge worker unemployment rate nearing one-fifth. External advisors like Ben Bernanke participated in discussions on the new industries needed for high growth.
On the capital path, Anthropic has made its internal models and assumptions public, shifting resources from model training to policy dialogue and scenario tool deployment. The motivation is to shape public discussions on the economic impact of AI in advance and promote international rule-making to address potential rapid growth and unemployment risks.
Similar cases include past productivity predictions related to the internet and automation, as well as public discussions by labs like OpenAI on the economic impact of AI. Currently, there is an expansion phase from capability demonstrations to macroeconomic scenario modeling, with labs beginning to provide quantitative frameworks for policymakers to reference.
Essentially, this pertains to regulatory changes, where the mechanism is that when the speed of technological capability improvement exceeds the existing system's adaptability, high growth predictions force international coordination to establish "extraordinary rules." Otherwise, capital concentration and declining labor shares may exacerbate distribution conflicts, shifting the industry chain from pure technological competition to a parallel reconstruction of technology and governance.
ABAB News · Cognitive Law
The larger the growth figure, the more new industries are needed.
The more extreme the model, the more urgent the rule demands.
When technology accelerates discovery, institutions must accelerate in sync.