A new concern is spreading across U.S. stock markets: not only are employees at risk of job loss, but entire companies face potential obsolescence. While most economists believe fears of an AI-driven employment apocalypse are exaggerated, history shows that major technological breakthroughs have indeed triggered dramatic industry transformations. The information technology revolution of the 1990s led to a surge in productivity, fueling years of rapid U.S. economic growth. However, it also rendered many companies and even entire industries redundant—from travel agencies and stockbrokers to classified ads and newspapers, and video rental stores.
Economists anticipate that artificial intelligence will boost productivity, which is crucial for long-term economic growth. Yet investors are increasingly worried about the potential damage AI could inflict on capital and labor markets, especially since its disruptive impact may far exceed that of the internet boom. Anton Korinek, an AI expert at the University of Virginia, stated, "Is the scale larger this time? Yes—perhaps tenfold. The key difference from the 1990s is that the internet merely changed how information was disseminated, while AI is revolutionizing cognitive production. This affects a much broader segment of the economy."
It is certain that these are early speculations about a rapidly evolving and largely unproven technology, whose ultimate goal is to enhance worker productivity. Productivity essentially measures how much workers can output with existing tools; thus, when significant new tools like the internet or AI are invented, productivity tends to rise substantially. Data on U.S. productivity for the final three months of 2025 will be released later on Thursday. Economists generally avoid overinterpreting single-quarter figures due to volatility. Still, the overall trend has been upward. After significant fluctuations during the pandemic, U.S. productivity has grown at an average rate of 2.6% since early 2023—more than double the average rate of the decade ending in 2019.
There is vigorous debate about how much of this acceleration can be attributed to AI. However, even analysts who believe the new technology has not yet made major contributions mostly expect it to play a significant role soon. Improvements in labor productivity can enhance efficiency, enabling businesses and their employees to increase incomes without triggering inflation. Historically, economies have adapted to major technological breakthroughs—creating new industries and jobs previously unimaginable—thereby raising living standards.
Simon Johnson, a Nobel laureate economist at MIT, noted, "It is normal, perhaps even inevitable, for an industry to experience boom and bust cycles." But he warned that as companies fail, broader risks may emerge—especially if failing businesses have borrowed heavily. "The last thing you want is a shock to credit markets, and you absolutely don't want to involve the banking system."
For now, so-called "AI panic trading" remains minimal in U.S. capital markets. Since the launch of ChatGPT in November 2022, the S&P 500 has risen by about two-thirds. A significant portion of this gain has been driven by soaring valuations of AI companies and their suppliers, such as giants like Meta Platforms and Nvidia, though this brings risks if their technology disappoints.
Another set of factors—different from the above and behind recent market volatility—stems from the possibility that AI could deliver the anticipated productivity leap, or even exceed expectations. A little-known research firm, Citrini, outlined this view in a report that briefly sent the S&P 500 plunging early last week. Citrini's scenario of massive AI-driven white-collar job losses, set in 2028, is essentially science fiction. Currently, there is no evidence of such a trend, with U.S. unemployment at historic lows.
However, Daniel Keum of Columbia Business School, who studies how automation technologies like AI shift power dynamics within firms, believes signs of this transition are already visible. His research, based on comments from earnings calls and annual reports, finds that employers increasingly view employees as costs—one of many signs that power is tilting back toward management. Keum stated that even if companies are not yet cutting jobs or salaries, they are reducing benefits like healthcare, remote work options, and even free snacks. "These extras are the first things companies look at before considering lowering your wages."
When companies use technology to boost efficiency and reduce labor costs, it typically benefits corporate profits and shareholders. For example, Block, the fintech firm run by Twitter founder Jack Dorsey, announced on February 26 that it would cut nearly half its workforce, betting on AI-driven productivity gains. Its stock has since risen over 15%. But last week also provided an example of how productivity gains can negatively impact investors—involving the long-established IBM. Startup Anthropic claimed its AI tool could perform a task that previously required "teams of consultants": modernizing the obsolete programming language Cobol running on IBM computers. IBM's stock suffered its largest drop in 25 years, though it later recovered most of the loss.
Past tech booms have seen well-known companies decline—such as camera maker Eastman Kodak and video rental chain Blockbuster, both rendered obsolete by the internet. This aligns with what economist Joseph Schumpeter termed "creative destruction," the process that drives progress. Tom Barkin, President of the Richmond Fed, referenced this concept last week when asked if the Fed should counter AI's impact on businesses and labor markets. He said, "This has been happening in this country for hundreds of years. It's part of the essence of capitalism."
However, this may not reassure industries facing short-term risks, along with their investors and employees. Korinek listed vulnerable sectors, including back-office services, content creation, customer support, legal and financial analysis, and programming. He warned, "Ultimately, any company whose competitive edge relies on human expertise that AI can replicate will be impacted. The transition period carries risks of stranded assets, excessive debt burdens, and sharp market adjustments."
Comments