Nobel Economist Acemoglu Predicts AI Will Add Only 1.5% to GDP Over a Decade
In a forecast published by Microsoft, Nobel laureate Daron Acemoglu argues that artificial intelligence will deliver far smaller economic gains than many in the tech industry expect, with GDP growth of about 1.5 percent over ten years and at most five percent of jobs replaced.
Artificial intelligence will add only about 1.5 percent to global economic output over the next decade and will replace no more than five percent of jobs, according to a starkly cautious forecast from Nobel Prize-winning economist Daron Acemoglu that Microsoft has now published. The projection stands in sharp contrast to the bullish expectations that have driven hundreds of billions of dollars of investment in AI data centers, chips, and models.
Acemoglu's central argument is that the economic payoff from AI will not come from ever-larger language models or more powerful computing clusters. What matters, he contends, is whether the technology produces practical applications that fundamentally change how work is done. Without such applications, the gains remain limited to a narrow set of tasks and industries, leaving the broader economy largely untouched.
The forecast places Acemoglu, a professor at MIT and a recipient of the Nobel Memorial Prize in Economic Sciences, at odds with much of the technology industry. Major AI developers and their investors have promoted the idea that generative AI represents a general-purpose technology on the scale of electricity or the internet, capable of lifting productivity across the entire economy. Acemoglu's estimate of roughly 1.5 percent GDP growth over ten years is a fraction of what those scenarios imply.
His skepticism about job displacement is equally notable. The prediction that at most five percent of jobs will be replaced is far below the warnings of mass unemployment that have accompanied each wave of AI advancement. Acemoglu has previously argued that automation tends to reshape tasks rather than eliminate entire occupations, and that the pace of adoption is constrained by regulation, corporate inertia, and the difficulty of redesigning workflows around new tools.
The decision by Microsoft to publish the forecast is itself significant. The company is one of the largest investors in AI infrastructure and a leading partner of OpenAI, and its products increasingly embed generative AI features. Publishing a bearish economic assessment alongside its own commercial ambitions suggests that at least some parts of the industry are willing to entertain the possibility that the technology's economic impact will be slower and narrower than the prevailing narrative suggests.
For policymakers, the forecast carries implications that extend beyond investment strategy. If AI delivers only modest productivity gains, the case for sweeping deregulation or massive public subsidies framed around an imminent economic transformation becomes harder to sustain. At the same time, a slower transition would give governments, educators, and employers more time to prepare workers for the changes that do occur.
The debate over AI's economic value has intensified as spending on the technology has soared. Data center construction, energy demand, and semiconductor supply chains have all been reshaped by the expectation of rapid returns. Acemoglu's analysis suggests that those returns may take much longer to materialize, and that the most important breakthroughs will be in unglamorous applications that quietly change how businesses operate rather than in headline-grabbing model releases.
Whether his forecast proves accurate will depend on factors that are difficult to predict, including the pace of innovation in areas such as robotics, healthcare, and scientific research. But the publication of his views by a company at the center of the AI boom ensures that the more cautious case will reach an audience that might otherwise hear only the most optimistic projections.
4
