In the industries most exposed to the impact of artificial intelligence, not only productivity increases, but also jobs and wage growth

Impact forecasts c artificial intelligence range from Art apocalyptic yes utopian. Report for October 2025 from the Democrats in the Senate, for example, predicted AI will destroy millions of US jobs. a couple of years ago McKinsey consultant forecast Artificial intelligence will add trillions to the global economy, while highlighting that job losses can be mitigated by teaching workers how to do new things.

The problem is that many of these claims are based on predictions, overly simplistic surveys, or thought experiments, rather than observable changes in the economy. Because of this, it is difficult for the public, and often politicians, to know who to trust.

As a a labor economist who is a teacher As technology and organizational change affect productivity and well-being, I think it’s best to start with actual data on output, employment and wages – all of which look relatively more encouraging.

AI and Jobs

In one of mine new scientific works with economist Andrew Johnstonwe examined how the impact of generative artificial intelligence affected industries across America between 2017 and 2024, using administrative data covering nearly all employers. Our analysis covered the responsible period when the use of generative AI has explodedwhich allows us to analyze the effect across businesses and industries.

We measured the impact of artificial intelligence using task-level data according to each industry and professional composition of the state’s workforce before the pandemic. A state and industry with more workers performing language processing, coding or data processing tasks received a higher impact score, for example, compared to a state with more plumbers and electricians.

We then took it exposure rating by occupation and looked at changes in the standard deviation of occupational exposure, comparing it to labor market and GDP by state and industry from 2017 to 2024.

Think of the standard deviation as the gap between a paramedic, whose job focuses on physical assessment, emergency response, and hands-on care that AI can’t easily replicate, and a public relations manager, whose job involves crafting messages, analyzing sentiment, and synthesizing information that AI tools are good at handling. This AI impact divide is roughly what we’re measuring when we ask: Does being on the higher impact side of the divide change the trajectory of your industry?

This data allowed us to answer two questions: When AI tools became widely available after ChatGPT’s public release in late 2022, did states and industries that were more exposed to generative AI become more productive, and what happened to workers?

Our answers are more encouraging and more nuanced than most public debates suggest.

We found that industries in states that were more exposed to AI experienced faster productivity growth starting in 2021—before ChatGPT became available to the public—thanks to enterprise tools already built into professional workflows, including GitHub Copilot for software development, Jasper for marketing and content writing, and Microsoft’s GPT-3-based business software. For example, in 2024, industries with one standard deviation higher exposure to AI saw a 10% increase in productivity, 3.9% in jobs, and 4.8% in wages than comparable industries in the same state.

These patterns suggest that, so far at least, AI has acted as a productivity-enhancing tool that increases employment and wages, rather than simply replacing work.

Magnification vs displacement

An important distinction in the data is between tasks where AI works with humans and tasks where AI can act more independently. In sectors where AI primarily complements workers (such as marketing, writing, or financial analysis), our data show that employment increased by about 3.6% per standard deviation of increased exposure.

In sectors where AI can perform tasks more autonomously, including basic data processing, boilerplate code generation, or handling standardized customer interactions, we found no significant employment changes, although wage growth slowed for workers in these roles.

These findings suggest that when artificial intelligence lowers the cost of a task and increases worker productivity, companies increase production enough to increase demand for labor in general—the same logic that explains why power tools haven’t eliminated construction workers.

The economic question is not whether any task will disappear. It depends on whether businesses and workers can reorganize quickly enough to create new productive combinations. And so far, in most sectors, our data suggests that they can.

But state policies also matter: These benefits have been concentrated in states with more efficient labor markets, meaning that the impact of generative AI on workers and the economy also depends on the types of policies and institutions in the local economy.

It is important to note that these findings are not limited to occupational exposure. In additional work with co-authors from the Bureau of Economic Analysis, we found similar effects on GDP and employment when looking at actual use of AI—that is, how often workers use AI. Picture on Gallup working groupwe measured workers who actively use AI daily or several times a week. We found that each percentage point increase in the share of frequent AI users in a state and industry is associated with about a 0.1% to 0.2% increase in real output and a 0.2% to 0.4% increase in employment.

To put this into context: The share of frequent users of artificial intelligence across all occupations rose from about 12% in mid-2024 to 26% by the end of 2025, a shift we estimate corresponds to about 1.4% to 2.8% higher real output – or about 1-2 percentage points of annual growth over that period.

New technologies rarely leave work untouched. But they also rarely eliminate the need for human input entirely. Instead, they are changing the composition of work, as our research shows. Some tasks are shortened. Others are expanding. New ones are emerging that were previously too expensive or too difficult to do at scale. Simply put, some classes can get away, but most of them just change.

Anyway, the trends are documented here is likely to strengthen rather than disappear. Not only are generative AI tools rapidly improving, but the experimentation, research, and development that many workers and companies are engaged in could pay big dividends. These investments, often referred to as intangible capital,are usually unlocked a few years after the technology enters the scene, once additional investments have been made.

The role of companies and managers

Whether AI will lead to worker anxiety or adaptation depends in part on what happens inside the organization. Using additional data In the 2026 article, compiled over the years by the Gallup Workforce Panel, which covers more than 30,000 US employees from 2023 to 2026, I found that the adoption of generative artificial intelligence in the workplace grew rapidly during this period, with the percentage of workers using artificial intelligence often increasing from 9% to 26%.

But the more important finding is that adoption was much more prevalent where employees believed their organization had communicated a clear AI strategy and where employees said they trusted management. This suggests that increasing the adoption and effective use of artificial intelligence depends not only on the availability of the technology, but also on whether managers make its use understandable, reliable and safe.

Where such clarity exists, frequent use of AI is associated with greater engagement and job satisfaction, and even reverses the burnout penalties seen elsewhere.

In other words, AI’s broader economic impact depends not only on how advanced the tools are, but also on whether companies and managers create an environment where employees can experiment, reorganize tasks, and integrate new tools into productive routines. That is, if employees do not feel psychologically safe to experiment, they will be less likely to use AI, and especially they will be less likely to use it for more valuable work.

I believe that it is this adaptation that makes labor markets more resilient than the most alarming forecasts suggest.

This article is reprinted from Conversation under a Creative Commons license. Read it original article.