AI will lead to mass unemployment, a new paper says - but other researchers are skeptical

By Jules Rimmer

'Thought-provoking' but 'implausible' is how Evercore analysts regard the new paper from Citrini Research

The Citrini Research paper imagines a scenario in which widespread AI adoption leads to mass unemployment. That would only come to pass if the government did nothing, Evercore said.

The publication of a paper from Citrini Research this week has caused as much share-price destruction as it has generated controversy, with multiple stocks and sectors impacted by the dystopian threat of artificial-intelligence disruption and societal chaos it posited.

But according to a research report published Tuesday by the investment bank Evercore, there's just one problem: This scenario isn't going to happen.

Evercore's critique of the paper highlights three obstacles to the realization of the Citrini scenario. First, human engagement may prove to be critical in a large number of functions, while non-AI businesses may prove more resistant to competition. Second, AI may run into cognitive limits that can't be solved by compute, or restrictions on energy use that prevent full implementation. Third, the scenario involves many assumptions about a lack of central policy response, no economic benefits and no "Schumpeterian creative destruction."

"AI agents handle many-weeks-long research and development tasks. The exponential steamrolled our conceptions of what was possible," Citrini Research said in its paper.

By "Schumpeterian creative destruction," Evercore authors Krishna Guha, Marco Casiraghi and Gang Lyu refer to the continuous destruction of older occupations and industries that are superseded by more modern, innovative and productive models.

The Evercore riposte acknowledges that there is a possibility of "extraordinarily rapid job and non-AI wealth destruction" that could again usher in a fresh era of financial distress and a world of zero-interest rates and ultrastimulative fiscal policy. The note also concedes that the scenario projected by Citrini's researchers is "thought-provoking" and throws up conditions and circumstances that do merit study and discussion.

One of the objections cited by the Evercore team is to Citrini's suggested scenario in which the vast wealth that accrues to owners of AI and compute fails to stimulate either new consumption or additional investment demand. The team believes this notion of "satiated consumption" is implausible. New services and activities will undoubtedly develop to match the consumption such new wealth would generate, they say.

Referring to the K-shaped economy, the Evercore authors also opine that blue-collar workers would be less affected because robots will not be able to fully replicate manual activities. They reckon blue-collar productivity wouldn't rise anywhere near as fast as in cognitive industries, and therefore those workers would see large relative wage gains.

Evercore also asserts the Citrini scenario confuses the individual company implications and the macroeconomic implications of agentic price discovery. They hold up as an example how credit-card companies may be disintermediated and see margins contract, but merchants and consumers would benefit by this process of eliminating the middleman, allowing economic activity to actually increase. Real-estate agents, too, might see their margins on home transactions collapse, but homeowners would benefit and housing activity might even see an uptick. Evercore calls this process "the erosion of oligopoly rents."

The note's authors draw a fascinating analogy to the Citrini paper's prediction of massive job losses with the Marxist notion that capitalism would "ultimately destroy itself by immiserating the petit bourgeois and working class until it had no consumers left." AI may begin by destroying more jobs than it creates, Evercore concedes, but that's unlikely to be the end of the story.

The final point the Evercore appraisal makes is that the Citrini forecasts don't seem to make allowances for a monetary or fiscal response in the event that unemployment, especially among politically influential white-collar workers, is driven to 10% and beyond. Evercore speculates that taxation might suddenly shift from labor to compute or wealth, and a return to zero-bound interest rates would follow.

-Jules Rimmer

This content was created by MarketWatch, which is operated by Dow Jones & Co. MarketWatch is published independently from Dow Jones Newswires and The Wall Street Journal.

(END) Dow Jones Newswires

02-24-26 1059ET