The surprising reason why women are using AI less often than men

A decade ago, when Claire Burgi moved to New York City, she decided to cut meat out of her diet. The 33-year-old actor and audiobook narrator, who lives in Queens, grew up in California, where she’d seen the effects of climate change firsthand. She knew that meat consumption was a major driver of greenhouse gas emissions and that vegetarianism was a way to help conserve resources and reduce pollution.

“When I was young, it rained a lot,” she says. “Now, it rains much less. All the fires are astoundingly horrific.” The December 2017 Thomas wildfire burned more than 280,000 acres in and around Burgi’s hometown of Ventura, just north of Los Angeles. 

“I just didn’t want to be contributing to anything that was causing that,” Burgi says. She recently made another major decision to reduce her eco-footprint: not to use generative AI.

She’s been shocked, she says, by research showing how much electricity that the underlying technology generative AI tools like ChatGPT use, and how much this is raising carbon emissions. One paper published in 2023 predicted that AI-related infrastructure would soon consume six times more water than is used in Denmark yearly. Another piece of research from 2024 showed that a request made through ChatGPT consumes 10 times the electricity of a Google search.

“They make me think of Frankenstein,” Burgi says of AI models. There have been times in history, she says, when humans have “acted without any idea of what the consequences would be, because it was convenient for us in that moment.” 

Right now, she adds, “that’s what’s happening with AI.”

In general, women have been slower to adopt AI use than men. This gender gap has been well documented over the last few years. According to Harvard Business School associate professor Rembrand Koning—who synthesized data from 18 studies covering more than 140,000 individuals across the world—women are about 20% less likely than men to directly engage with this new technology. 

What’s less clearly established is the precise reason why. But when it comes to environmentally motivated reasons, Burgi isn’t alone.

“Environmental angst”

The reasons for this gender gap vary. Some studies indicate that women are less likely to trust that gen AI providers will keep their data secure. Other research shows that women are more fearful of a loss of control that comes with these technologies—which is, for example, reflected in their more muted enthusiasm for driverless cars. Studies have also shown that women are more likely to avoid AI because of fears that it could steal their job, and still other studies have found that women are more concerned than men about the ethical implications of AI use. 

But a growing body of research also indicates that a sizable chunk of the gap might be attributable to the type of environmental angst that people like Burgi feel. Earlier this month, academics at the University of Oxford published a paper showing that the reasons for the adoption gap are manifold, but that environmental concerns certainly play a significant part. 

The research, titled “Women Worry, Men Adopt: How Gendered Perceptions Shape the Use of Generative AI,” draws on survey responses from 8,000 individuals in 2023 and 2024 across the U.K. It established that 14.7% of women and 20% of men reported using gen AI tools frequently—at least once a week—in a personal context. This corresponds to a gap of just over 5 percentage points. 

But the gap widens significantly in subsets of respondents who admit to being concerned about environmental and mental health risks. Among those who say they are worried about the climate, the gender gap is 9.3 percentage points; for those concerned about the mental health impact of these new technologies, it widens to 16.8 percentage points. Among older users of artificial intelligence, the gender gap for concerns about AI’s climate effect is particularly wide: almost 18 percentage points. 

These findings echo previous research showing that women are more likely to display “eco-anxiety” than men—a phrase that’s been coined to refer to the mental health distress caused by climate change, ranging from concerns about the impact of extreme weather to the future of biodiversity. And the academics at Oxford write that their findings align “with evidence of greater social compassion and moral sensitivity among women.”

Counterintuitive findings

Fabian Stephany, a departmental research lecturer at the Oxford Internet Institute and one of the authors of the study, says that one of the most interesting things his research found was that some common preconceptions about AI usage weren’t corroborated. 

There’s a widely held assumption, for example, that greater tech literacy translates into higher use and adoption. But he found that this isn’t always the case. In fact, in some cases, greater literacy and knowledge about AI actually drove down use. 

Also of note, the research found that among those who said they were concerned about AI’s impact on the environment and on mental health, women’s concerns were more likely to translate into action than men’s concerns. In other words: Women were more likely to stop using AI because of the way they felt about it. 

Asked why that might be, Stephany said he could only speculate. Research done by academics in Iran in 2022, though, might provide an answer: It shows that women generally lean toward a more collectivist mindset—reflected in concerns for society, for example—while men tend to lean toward a more individualist one.

Women are “the canary in the coal mine”

Stephany says that the last thing he wants people to take away from this research is that women need to change or “be fixed” in some way. “Their concerns are real,” he says. 

“Women are like the canary in the coal mine. And we should listen to these concerns,” he adds. “The important thing isn’t to tell women to be more optimistic—it’s to address the harms.” And, he says, these harms can be addressed. 

“Biases can be reduced, carbon footprints can be lowered, smaller models can be run locally,” Stephani says. “We don’t have to wait for future breakthroughs. We can reduce harms now.”

His research suggests that “there’s a sizable market of people with strong convictions about AI consumption,” he says. “A more green, sustainable, inclusive gen AI model would have a clear target market.”

And there are already platforms available that seem to be tapping into this market that Stephany mentions. GreenPT, for example, bills itself as a platform that runs on renewable energy and is hosted in Europe “for strict data protection.” Viro, another platform, funds clean energy projects and markets itself as a “climate-neutral” alternative to less environmentally conscious options.

Speaking at the Massachusetts Institute of Technology last year, Sam Altman, the CEO of OpenAI, which operates ChatGPT, tried to allay fears that his technology might be accelerating climate change by framing it as a tool to enhance sustainability. “If we have to spend even 1% of the world’s electricity training powerful AI, and that AI helps us figure out how to get to non-carbon-based energy or to do carbon capture better, that would be a massive win,” he said.

As for Burgi, she would want to see a lot of changes before even entertaining the idea of intentionally incorporating AI use into her daily life. She doesn’t think that anything could meaningfully allay her ethical concerns about AI. “Especially as an artist, I just don’t feel morally aligned with using AI,” she says. 

In terms of her environmental concerns, she’s similarly skeptical. 

“If there was more transparency, and if it seemed like more thought and care was being put into these things—if it wasn’t just about greed and capitalism—then I might consider it from an environmental concern,” she says. “But right now? I don’t really see any of that happening.”

View the full article