AI’s Job Impact: Gains Outpace Losses

It seems like pretty much all we hear about artificial intelligence (AI) is that it is destroying jobs. At least in 2024, that was not the case.

In 2024, AI growth generated thousands of jobs, with estimates of more than 8,900 employees added to the U.S. economy to develop, train, and operate AI models, including machine learning engineers and data scientists.

On top of that, AI firms’ expansion of data centers fueled a surge in construction activity. Each large-scale data center requires roughly 1,500 on-site workers and can take up to three years to complete. Given the record volume of data centers being built, this translated into over 110,000 construction jobs in 2024. Some estimates also suggest that data centers have a strong local multiplier effect, generating an additional 3.5 jobs for every one job inside the data center.

Altogether, AI created about 119,900 direct jobs last year.

In contrast, outplacement firm Challenger, Gray, & Christmas estimates that approximately 12,700 jobs were lost due to AI in 2024, far less than the number created by the technology. Additionally, that figure represents just 0.1 percent of all layoffs last year. (See Figure 1).

Figure 1: U.S. jobs lost and gained from AI (2024, in thousands)

While headlines warn that AI will automate large numbers of jobs and trigger widespread layoffs, the reality is far more grounded. A recent MIT report estimates that 11.7 percent of the labor market could, in principle, be automated. But whether automation will reach that level remains uncertain, and history shows that automation tends to reallocate work rather than eliminating it outright.

If the past 30 years are any indication, fears of mass displacement are misplaced. Job losses as a share of total employment have trended downward even as automation and digital tools have become more widespread.

Taken together, the evidence tells a clear story: The employment gains from AI and the data center buildout dwarf the displacement effects from automation. Instead of hollowing out the workforce, AI is reshaping it, creating new job opportunities across the economy.

Finally, two broader points are worth emphasizing.

First, technology has no impact on overall employment in the moderate to long term—for example, after an adjustment period of roughly two years. The United States has consistently maintained full employment even in the face of technological disruption, setting aside cyclical downturns.

Second, we cannot lose sight of productivity. The only sustainable way to boost worker incomes is through technological innovation, and AI holds significant promise on that front.

So long as policymakers and the public avoid job panics, and the counterproductive policies they tend to generate, AI is likely to continue boosting productivity and supporting steady, broad-based improvements across the economy.