The biggest AI disruption may not be mass unemployment. It may be a widening economic divide between men and women.
For years, the artificial intelligence debate has revolved around one terrifying question: Will AI take our jobs?
But new research from Pew Research Center suggests we may be asking the wrong question.
The more urgent question is: Who is benefiting from the AI economy, and who is being left behind?
The early evidence reveals a troubling pattern. Employment is growing in occupations most exposed to artificial intelligence. But men are capturing substantially more of that growth than women.
And the divide becomes especially severe for workers without a college degree.
The Numbers Should Make Everyone Pay Attention
Pew researchers analyzed 480 occupations, comparing employment before the generative AI boom, from 2017 to 2019, with employment between 2023 and June 2026.
Their finding was striking.
For every one-point increase in an occupation's AI exposure score, men's employment growth was 6.3%, compared with just 0.2% for women.
In other words, as occupations become more exposed to AI, the employment growth advantage increasingly shifts toward men.
This does not mean AI caused that disparity. Other economic forces may be contributing. But the pattern raises an uncomfortable question about who is positioned to prosper as technology transforms the workplace.
The Real Divide Is Education
The most alarming findings concern workers with a high school diploma or less.
For every one-point increase in AI exposure, men's employment grew by 8.8%, while women's employment declined by 9.5%.
That is an 18.3-percentage-point difference in employment growth.
Among workers with some college education, the gender gap was smaller but remained significant.
Among workers with bachelor's degrees or higher, however, researchers found no statistically significant relationship between AI exposure and employment growth for either gender.
The implication is hard to ignore: The intersection of gender, education and technology may become one of the defining economic fault lines of the AI era.
The Invisible Vulnerability of Administrative Work
Consider where many women without four-year degrees work.
Reception. Customer service. Secretarial support. Administrative operations.
These jobs frequently involve processing information, answering questions, scheduling, preparing documents and coordinating routine activities.
They also involve exactly the kinds of tasks generative AI can increasingly assist with or automate.
Meanwhile, many men without college degrees work in construction, transportation, freight handling and other physically intensive occupations that remain less exposed to current AI capabilities.
That distinction matters.
The vulnerability is not necessarily that women are less capable of using AI. It is that the occupations employing many women without degrees overlap heavily with the tasks AI is being designed to perform.
And when technology changes the economics of those tasks, the consequences can be deeply unequal.
This Is Not Just a Technology Problem. It Is an Opportunity Problem.
AI promises extraordinary productivity gains.
But productivity gains do not automatically translate into broadly shared prosperity.
A company can become more efficient while reducing entry-level opportunities. An industry can expand while some workers lose bargaining power. Employment can grow overall while particular groups struggle to participate.
The Pew findings do not establish that AI is responsible for the employment gap. Yet they offer an early warning that technological progress and equitable economic progress are not the same thing.
If businesses and policymakers focus exclusively on how many jobs AI creates or eliminates, they risk overlooking a crucial question: Who has access to the opportunities that survive and emerge?
The Next AI Divide Could Be Bigger Than the Digital Divide
The response cannot simply be to tell everyone to learn AI.
Workers need affordable, practical training. Employers need to create pathways from vulnerable administrative roles into higher-value work. And workforce policies must account for the people least able to absorb another round of economic disruption.
Above all, AI adoption needs to be measured not just through corporate productivity and shareholder returns, but through wages, mobility and opportunity across education levels and demographic groups.
Because a technological revolution that expands opportunity for some while narrowing it for others is not universally shared progress.
The Bottom Line
The most important AI employment story may not be about humans versus machines.
It may be about which humans get ahead.
The early numbers suggest men, particularly in the comparison of AI-exposed occupations, are experiencing stronger employment growth than women. And among workers without college degrees, the divide is especially pronounced.
We should not mistake correlation for causation. But we should not ignore an emerging inequality simply because the technology is still young.
The AI revolution is underway. The question is no longer just how powerful the technology will become. It is who will share in the prosperity it creates.
And if we get that wrong, the greatest cost of AI may not be the jobs that disappear.
It may be the opportunities that millions of people never get.
What do you think? Will AI ultimately narrow economic inequality, or accelerate the divides already shaping our workforce?
Source: Pew Research Center, “Men Are Experiencing More Job Growth Than Women in AI-Exposed Occupations,” October 22, 2026.
