AI Boosting Jobs and Wages, Not Replacing Workers

The fear that AI will steal millions of jobs just hit a major roadblock: the actual data tells a completely different story.

Labor economist Christos Makridis and colleague Andrew Johnston analyzed employment records covering nearly every US employer from 2017 to 2024. They tracked what happened when generative AI tools exploded into workplaces after ChatGPT's 2022 launch.

The results flip the doom-and-gloom predictions on their head. Industries with high AI exposure didn't shed jobs. They gained them.

In 2024, industries one standard deviation higher in AI exposure saw productivity jump 10% compared to similar industries. More surprising: those same industries added 3.9% more jobs and paid 4.8% higher wages.

Think about the difference between a paramedic handling physical emergencies and a PR manager drafting communications. That gap represents roughly one standard deviation in AI exposure. The PR manager's industry, more exposed to AI, actually grew faster.

The growth started earlier than most people realize. Productivity gains began appearing in 2021, before ChatGPT went public, driven by enterprise tools already working behind the scenes. Software developers used GitHub Copilot. Marketing teams relied on Jasper. Business teams integrated GPT-3 powered applications.

The pattern held when researchers measured actual AI use, not just potential exposure. Using Gallup data, they found states and industries where workers actively used AI daily saw 0.1% to 0.2% higher output per percentage point of frequent users.

Between mid-2024 and late 2025, frequent AI users jumped from 12% to 26% of workers. That shift translated to roughly 1.4% to 2.8% higher real output across those sectors.

The key distinction lies in how AI integrates into work. In fields where AI complements human skills like marketing, writing, and financial analysis, employment rose 3.6% per standard deviation of exposure. Workers became more productive, companies expanded, and demand for labor increased.

Even in sectors where AI handles tasks more independently like basic data processing or standardized customer service, employment stayed stable. Workers in these roles saw slower wage growth but kept their jobs.

The pattern mirrors what happened with power tools in construction. When technology makes tasks cheaper and faster, companies typically expand output enough to need more workers, not fewer.

The Bright Side

This isn't just theoretical hope. It's happening right now in real businesses across America. The same AI tools that sparked fears of mass unemployment are actually helping companies grow fast enough to hire more people and pay them better.

State policies matter too. The benefits concentrated in states with more efficient labor markets, showing that smart local policies can amplify AI's positive effects on workers.

The data reveals what headlines often miss: when humans and AI work together, both productivity and prosperity can rise.