‘Reese is right’: The AI skills gender gap is real

When Reese Witherspoon urged women to learn artificial intelligence, the response was immediate. Critics accused the actor and entrepreneur of putting the burden on women to solve a problem created by tech companies and employers. Others pointed to AI’s environmental costs, ethical concerns, and growing threat to creative work.

Witherspoon later clarified that she was not being paid to promote AI and said she doesn’t believe computers should replace humanity. She also acknowledged concerns about jobs and the broader social impact of the technology.

Still, beneath the backlash lies an uncomfortable question: was she wrong?

“It’s actually a horribly accurate concern,” said Dr. Mary Noble-Tolla, Director of Research and Content at Lean In, referring to the risk women face in the AI transition. “If women don’t really accelerate their learning of AI, then all of the gender gaps that we definitely see will start to get worse, not better.”

A growing body of research supports this concern and suggests that women face a double risk in the AI era. They’re more likely to work in roles being reshaped by generative AI, yet less likely than men to use AI tools regularly at work.

Last year, the United Nations’ International Labour Organization (ILO) reported that women in high-income countries are nearly three times more likely than men to work in jobs with the highest exposure to generative AI automation. It said that 9.6% of female employment falls into the top-risk category, compared with 3.5% for men, underscoring how unevenly disruption could be felt as AI spreads through the workforce.

In the U.S., recent research from Brookings found that 6.1 million workers are in occupations with both high AI exposure and low capacity to adapt to job displacement. Of those workers, 86% are women, with many concentrated in clerical and administrative roles.

At the same time, Lean In research reported that men are 22% more likely to use AI daily at work, making them more likely to become early “power users” of the technology.

That combination could have lasting consequences for pay, promotions, and who gets ahead.

“We can be clear-sighted about the risk, but I think just pragmatically, Reese is right,” Noble-Tolla said.

Why women face a double hit

The issue isn’t that women are less capable or less interested in the technology; it’s where women work, how new tools are introduced, and who is encouraged to experiment first.

Noble-Tolla shared that women remain overrepresented in administrative support, clerical, customer service, and coordination-heavy roles, many of which involve routine cognitive tasks that AI can now perform faster and more cheaply.

“I just feel like all of these things are a perfect storm,” Noble-Tolla said. “Women are more likely to be in roles that can be displaced, and they’re less likely to be building AI skills day to day.”

Ariane Hegewisch, Senior Research Fellow at the Institute for Women’s Policy Research, said the picture is more nuanced than headlines suggest. Women, she noted, are hardly strangers to workplace technology, with many already working extensively with software, systems, and digital tools.

“A lot of the work that women do is on tech,” Hegewisch said, arguing the conversation should focus less on whether women need to “catch up” and more on how AI is applied in real workplaces.

She added that organizations often get better outcomes when employees are involved in deciding how AI can improve work rather than simply replace it.

Lean In’s findings also suggest that confidence and culture matter. Entry-level men were more than 50% more likely than women to say their manager had encouraged them to use AI, according to Noble-Tolla.

“Women are viewed as less competent than men,” she said. “They're not being encouraged in the same way.”

That matters because AI fluency is built through repetition and experimentation. Employees who are nudged to test tools, automate tasks, and learn by doing can gain an advantage over those who are more hesitant to embrace the technology.

There are also trust issues, according to Noble-Tolla. She said women often express greater concern about privacy, bias, ethics, and being judged as less competent or “cheating” for using AI.

READ MORE: Workers Are Using A.I. They Don't Trust

Closing the AI gender gap

Telling women to simply “learn AI” misses the bigger issue. Hegewisch said the more important question is how organizations and HR leaders introduce new technology and whether employees are included in shaping how it’s used.

“If you involve employees and collectively discuss how the new technology or AI can be used and how processes can be redesigned, you are likely to get more productive outcomes,” Hegewisch said.

Noble-Tolla added that organizations should begin by collecting internal data to understand whether a gender usage gap exists and why. A short employee survey, she suggested, can reveal whether women’s concerns about AI are about privacy, fairness, lack of training, or limited access to tools.

Noble-Tolla said training should move beyond basic introductions and be tied to employees’ actual responsibilities, helping them use AI to improve workflows and repetitive tasks. She also said companies need to ensure equitable access to AI tools so workers in administrative and support roles are not left behind as adoption accelerates.

Leadership visibility can also make a difference. Noble-Tolla said women may feel more confident embracing AI if they saw more female voices shaping the technology itself and raising ethical or safety concerns from positions of influence.

READ MORE: Women leaders take on 'active strategic roles' amid AI adoption

She noted that AI leadership remains heavily male-dominated, making representation in the next phase of workplace change especially important.

“This is a really big technological change, and it’s not going to stop or go away. We can’t stop it,” Noble-Tolla said. “The best thing we can do realistically is try and upskill, try and learn as much as possible so that we understand what’s going on and can put ourselves in positions where we have influence.”