A new peer-reviewed economics paper from researchers at the University of Pennsylvania’s Wharton School and Boston University has delivered one of the starkest warnings yet about artificial intelligence’s long-term economic impact: left unchecked, AI-driven automation could systematically collapse consumer capitalism itself.
Published in March under the title The AI Layoff Trap, economists Brett Hemenway Falk and Gerry Tsoukalas argue that competitive market incentives will push firms into an automation race so extreme that they eventually eliminate the very consumer demand their businesses depend on. Their conclusion is blunt: “At the limit, firms automate their way to boundless productivity and zero demand.” ()
The model outlines a dangerous feedback loop. Individual companies that replace workers with AI reduce labour costs and gain short-term efficiency. But when this strategy is repeated across an entire economy, displaced workers lose wages and cut spending. Since workers are also customers, aggregate demand begins to shrink.
As consumer spending falls, firms respond rationally by cutting costs even further — usually through more automation. The cycle then accelerates: layoffs reduce demand, falling demand encourages more layoffs, and the process becomes self-reinforcing.
In effect, every firm behaves logically in isolation while collectively driving the economy toward systemic failure.
The paper suggests this dynamic is not simply a labour market problem but a structural flaw in competitive capitalism under rapid AI adoption. More capable AI and more intense market competition both worsen the outcome, according to the researchers.
Notably, Falk and Tsoukalas tested several commonly proposed policy solutions, including universal basic income, retraining programmes, worker ownership models, capital income taxes and corporate coordination. None were sufficient to prevent the collapse in their model.
The only effective intervention was a Pigouvian automation tax — a levy on each AI-driven labour replacement that forces firms to internalise the broader economic damage caused by shrinking consumer demand. In other words, companies would have to pay for destroying purchasing power before automating jobs away. ()
The warning arrives amid already mounting concerns over AI-linked redundancies. Major technology firms have continued large-scale layoffs through 2025 and early 2026 while aggressively investing in generative AI and autonomous systems.
The broader implication is unsettling: no malicious actors are required for economic collapse. Rational firms, responding correctly to competitive incentives, may be enough.
For decades, economists assumed technological disruption would eventually create new industries and jobs to offset displacement. The AI Layoff Trap challenges that optimism by arguing AI may outpace labour market adaptation on an unprecedented scale.
If the paper’s mathematics hold, the future threat may not be runaway AI itself — but a market system automating so efficiently that it destroys its own customer base.