US Data Links AI Adoption With Productivity Gains and Professional Job Losses

AI’s labor-market effects are starting to appear in US data, but the early signal is narrow: industries that use AI more heavily have recorded faster productivity growth alongside signs of weaker employment. That combination deserves attention, but it doesn’t establish that AI caused the employment changes. Other economic forces may be affecting hiring at the same time, and the evidence described so far covers only selected industries in one country. The numbers point to a pattern worth tracking, not a settled account of what’s happening to workers.

Eric Basmajian, a US-based financial researcher, found that the information, professional services and finance sectors saw productivity growth accelerate from a 2.2 percent trend to 5.0 percent after 2022. In a social media post, he said AI-driven productivity gains were showing up in the jobs data, while estimating that only about 15 percent of jobs appeared affected so far. That estimate puts a useful boundary around the claim: even if the signal is real, it doesn’t describe a labor-market transformation reaching every occupation. The figures also don’t show how the affected jobs changed, or whether workers moved into other roles.

A purported internal Goldman Sachs report, circulated the following day, pointed to a growing relationship between AI exposure and unemployment in professional industries beyond technology. Finance, telecommunications, technical services and administration were among the fields cited. A relationship between exposure and unemployment can identify where to look more closely, but it doesn’t tell us by itself whether AI is the reason people are losing jobs. Hiring slowdowns, reorganizations and broader economic conditions can overlap with the adoption of new tools. The pattern suggests the labor impact may be spreading beyond software work, where generative AI has already become a common coding tool.

What the data can—and can’t—show

Productivity measures don’t reveal whether companies are producing better work, increasing revenue or simply doing the same amount with fewer employees. That distinction matters: a rise in measured productivity alongside falling employment could reflect workforce cuts rather than a broad improvement in output. It could also reflect changes in the mix of work being done or in how companies organize tasks. The figures cited here don’t settle those possibilities. Productivity can show that output per worker has changed; it can’t, on its own, explain why.

This is more significant than it looks, but the significance is limited. When productivity rises in sectors with higher AI exposure, it gives researchers and employers a reason to examine whether new tools are changing the amount of labor needed for particular tasks. It doesn’t prove that automation is replacing whole jobs, or that each productivity gain comes from AI. Similar patterns can have several causes, and separating them requires evidence that connects tool adoption to changes in work and hiring. (And this is the part most people overlook.) A sector-wide trend can mask very different outcomes across roles and employers.

The evidence is also specific to the US, where employment trends have been shaped by wider post-pandemic economic pressures. The Financial Times reported that comparable links between AI use, productivity growth and declining employment aren’t apparent elsewhere. That gap makes it harder to treat the US pattern as a universal result of AI adoption. Different labor markets and economic conditions may produce different signals, and the available account doesn’t explain which factors account for the difference. Research from Stanford has also examined how to distinguish AI’s effects from broader labor-market conditions in the post-pandemic jobs market.

Why exposure isn’t the same as displacement

AI exposure describes the potential for AI tools to affect work; it doesn’t establish that an employer has adopted those tools or cut staff because of them. A tool might speed up one task while leaving the rest of a job intact. It might also shift responsibilities without reducing the number of workers. For workers, those distinctions matter: the consequences could range from changes in daily tasks to reduced hiring or job losses, and the figures cited so far don’t show which outcome is occurring in each field.

If you’re working in a sector named in these findings, the practical signal is to pay attention to changes in both productivity and hiring, while treating claims about cause with care. A single measure can’t tell you whether a company is using AI to expand output, reduce costs or reorganize work. The evidence would be more persuasive if the same sector-level patterns continued and could be tied more clearly to actual adoption and employment decisions. For now, the numbers here are suggestive, not decisive.

What to watch next

For now, the figures offer a reason to watch AI exposure and hiring trends together, not proof of an economy-wide employment shock. The clearest test will be whether these sector-level patterns persist and appear in labor markets beyond the US. Researchers will also need to distinguish AI’s contribution from the wider forces shaping employment, a challenge highlighted by the Stanford work. The current evidence leaves those questions open.

That said, a narrow early signal can still matter. If productivity gains keep appearing alongside weaker employment in highly exposed fields, the relationship will deserve closer scrutiny. If the pattern fades or fails to show up elsewhere, that would change how broadly the current findings can be applied. Too soon to call.

More on AI and labor: Support for Unions Is Spiking as AI Threatens Jobs

The post There Are Now Growing Signs That AI Is Hurting Employment in the Job Market appeared first on Futurism.