AI REGS & RISKS | Is AI Going to Cause Economic Collapse?

What happens to the economy when AI-powered machines begin replacing large portions of human cognitive work? Some analysts believe we may be approaching the most disruptive economic shift since the Industrial Revolution. Others believe AI will produce the opposite outcome: the largest productivity boom in modern history. Both scenarios are plausible. Right now, the global economy is trying to understand what happens when intelligence itself becomes a scalable resource.

The AI Shock is Happening Faster than Ever

In just the past year, AI models have made dramatic advances in reasoning, long-context processing, and autonomous agents capable of completing complex multi-step tasks as well as, or better than, humans. Historically, technological revolutions unfolded gradually. Electricity, railroads, and the internet all took years or decades to diffuse through the economy. AI is moving far faster. The reason is simple: unlike previous technologies that primarily automated physical labor, AI directly targets knowledge workers.

The world of software development is already providing an early bellwether. Productivity gains that once took months or years to materialize are now appearing in weeks or days. Spotify’s co-CEO Gustav Söderström recently revealed during a February 2026 earnings call that some of their most senior developers had not written a single line of code since December 2025.

That raises an obvious economic question. What happens when a technology begins performing the tasks that power the modern knowledge economy?

Doomsday View: The Economic Collapse Scenario

A recent research report titled “The 2028 Global Intelligence Crisis,” published by Citrini Research, outlines a hypothetical scenario for the year 2028 in which rapid adoption of agentic AI triggers an economic downturn. The report argues that AI could cause a severe macroeconomic disruption by replacing large numbers of white-collar workers. Wage income would fall, consumer spending would decline, and corporate revenue would drop. Companies would then respond by cutting more workers, creating a self-reinforcing negative economic loop.

What makes this scenario particularly unsettling is the scale of potential disruption. Previous technological revolutions often displaced specific industries. AI, by contrast, targets a wide range of knowledge work simultaneously: software development, finance, legal services, marketing, customer support, research, and more. White-collar professionals also tend to drive major areas of economic activity, including housing, discretionary spending, and financial investment. A rapid reduction in these incomes could ripple through the broader economy.

It is not surprising that this narrative has gained traction on Wall Street and in economic circles. But it rests on one very large assumption: that demand stays fixed.

Expansion: The Productivity Argument

Many economists argue that the collapse thesis misunderstands how productivity shocks historically affect economies. When the cost of producing something collapses, demand rarely stays the same. Instead, consumption expands.

Cheap electricity did not reduce the demand for appliances. It created entire industries around them. Cheap computing did not eliminate software jobs. It produced an explosion of digital products and services. Cheap internet bandwidth did not shrink media consumption. It created entirely new forms of media.

AI could follow the same pattern. If the cost of producing software drops dramatically, the likely outcome is not fewer applications but exponentially more software. If the cost of producing content collapses, we may not see less content creation but vastly more.

The key insight is that productivity shocks tend to expand the economic frontier rather than shrink it. History suggests that when something becomes cheaper to produce, people simply find more ways to use it.

Labor Markets Don’t Vanish. They Restructure.

Historically, technological revolutions have restructured labor markets rather than eliminated them. Many areas of the economy remain difficult to automate. Skilled trades, healthcare, advanced manufacturing, hospitality, and physical services all require forms of dexterity, presence, or human interaction that AI struggles to replicate.

In many cases, AI may complement these roles rather than replace them. A technician augmented by AI diagnostics becomes more productive. A doctor using AI analysis can treat more patients.

At the same time, AI dramatically lowers the barriers to entrepreneurship. Tasks that once required entire departments — coding, marketing, accounting, and design — can be performed by individuals using AI tools. This could lead to an explosion of small businesses and independent creators.

The Keynes Prediction Problem

Economists have made similar predictions before.  In 1930, John Maynard Keynes famously suggested that productivity growth would eventually reduce the workweek to about fifteen hours. His logic seemed sound: if machines could produce more goods with less labor, people would work less.

Instead, something very different happened. As productivity increased, societies consumed more goods and services. New industries emerged. Expectations for living standards rose. Rather than working dramatically fewer hours, people simply found new ways to spend their time and income.

Human wants expand faster than productivity eliminates work. Technological progress changes what people do, but it rarely eliminates the need for economic activity altogether.

Not So Fast: The Real Unknown Is Adoption Speed

Perhaps the most important variable in the AI economy is not how powerful AI becomes, but how quickly it spreads. Technological revolutions do not transform the economy overnight. They diffuse gradually as organizations adopt new tools, restructure workflows, and retrain workers.

Early data suggests that while AI experimentation is widespread, deep organizational adoption remains uneven. Many companies are still in the early stages of integrating AI into their core operations.

This slower pace of adoption may actually stabilize the transition. If labor markets and institutions adapt gradually, the economy may absorb the disruption more easily than many fear.

The Bottom Line

Artificial intelligence may represent the most powerful productivity technology ever created. It could automate large portions of cognitive work and fundamentally reshape the knowledge economy.

But productivity revolutions have always looked destabilizing at the beginning. In many cases, they ultimately produced new industries, higher living standards, and entirely new forms of work.

We are still in the earliest stages of this transformation. The global economy is not collapsing. But it may be on the edge of the largest economic restructuring of the modern era.

Author’s Note

This article is written in memory of Cindy Taylor. Cindy was a remarkable person — a positive force of nature who brought energy, encouragement, and generosity to everything she did, right up to the very end.

I’m deeply grateful to her for the opportunity to write the AI Regs & Risk column — this being the 45th edition. Over the past couple of years we spent many hours discussing ideas and shaping articles together. As a new columnist, I benefited enormously from her thoughtful editorial guidance and steady encouragement.

One of my favorite memories is how excited she would get watching the readership grow — celebrating as the page views climbed from a few hundred, to a few thousand, and eventually well beyond that. She took genuine pride in helping writers find their voice and their audience.  And in keeping with Cindy’s insistence on transparency, I should also disclose that AI was used to help analyze source material and extract the key economic ideas discussed in this article.