Leaning In is not the answer for women not using generative AI
Folks need to stop blaming women for holding to their values and protecting themselves when it comes to using generative AI. We (all, not just us women) need to fix the root causes.
A recent HBS article reported a study concluding that women are using AI less than men and pointed out a few causes. [Note: they refer often to “AI”, but their focus is really generative AI, not ML and everything “AI” encompasses.]
“In many cases, the research suggests women are concerned about the ethics of using the tools and may fear they will be judged harshly in the workplace for relying on them”
A close reading of the study shows a nuanced summary that thoughtfully considers these and other systemic factors and recommends addressing them.
However, the bottom line of the article was basically to recommend that companies encourage women to ‘try harder’ to use AI so their careers won’t suffer.
This article & its title & overall attitude has been chafing me. “Leaning In” is not the answer. I’ll explain why and share what we ought to be doing instead.
What the Article and Study Said
No surprise to anyone who reads my writing here: I’m all for women (everyone) getting full support to try new tools, like genAI, where they make sense.
The HBS article mentions two systemic root causes of lower genAI use by women (bolded above): concerns about tool ethics, and fears of being judged harshly in the workplace for using AI.
Other root causes for lower use by women, pointed out by the full study, include the fact that current AI tools are trained on biased content.
However, in the conclusions of the article, the authors don’t push for actions to be taken on these root causes. They mostly encourage “addressing the usage gap”.
To me, that feels like blaming women and telling them to suck it up and just use unethical, biased AI tools anyway, or their careers will suffer. I’m not on board with any of that.
I do agree that it’s absolutely important for people of all genders and demographics to be able to participate fully in developing and shaping AI tools, the sooner the better. And the recommendations in the study are fine. What I don’t agree with is the advice in the article about how to achieve it.
The Real Answer for Two Root Causes
➡️ When AI-based tools aren’t ethical, forcing concerned women to override their moral compasses to use AI isn’t the answer. Getting the AI companies to prioritize bias mitigation and develop & deliver AI tools ethically is the answer.
Bonuses:
Everyone (including men who also care about ethics and biases) will get more ethical, less-biased tools to use, and
Everyone currently being exploited in how most AI tools are developed & used would be treated fairly. [For insights on exploitations and unfairness in AI tool development, see my March 11 article, “Top 5 things to know about AI ethics”.]
➡️ When women do hesitate more than men to use AI because they fear — with good reason! — that their competence will be judged more harshly than men’s, the answer isn’t (just) to tell women to use AI anyway and deal with the harsh judgments. The answer is to address the underlying biases that cause people of all genders to judge women more harshly than men for similar actions.
Bonuses:
Women will be treated more fairly in areas beyond AI too, and
Society will reap the benefits of greater involvement and contributions of women in our economy. [From the full study: “a persistent 25% usage gap could result in hundreds-of-billions of dollars of lost productivity and growth in the US alone”.]
I’m not saying these two root causes are easy to address. Clearly they’re not.
But that’s no excuse for mostly ignoring the root causes, and simply pushing women to compromise their values and get ready for more undeserved criticism. The full study report covers this, but the summary article glosses over it. We need to address these root causes along with other factors that may be holding women back from using AI. We’ll all be better for it.
Limitations of the Study
I also have questions on the relevance of the findings. The study covers 17 prior studies and adds an 18th new study based in Kenya. The report correctly points out that since “women tend to work in different types of firms, jobs, and occupations than men, they may be less exposed to this new technology.” Some of the 17 source studies did not compare AI attitudes among men and women within similar roles at similar experience levels, though.
Roles and experience, especially in male-dominated industries and firms, matter.
One of the summaries compares 5017 “US Working Professionals” found via Glassdoor. Absent information about the composition of the group, that’s unlikely to be a well-balanced group with respect to roles and levels of experience for US-based people of all genders.
In one study by BCG of 6558 “Tech Workers”, women were found to be more likely to use generative AI than men, specifically more senior technical women.
“The exception is a Boston Consulting Group (BCG) study showing that women are 3% more likely to use generative AI than men in a convenience sample of 6,558 tech employees (Barisano et al. (2024)). This positive gap is driven primarily by women working in senior technical roles, whereas junior women are significantly less likely to use generative AI than junior men (see Appendix A). This suggests that experience working in and using technology may play a critical role in eliminating the gender gap.”
Other Considerations
Aside from ethics, biases, and being judged harshly for using AI, there may be other underlying reasons why women are reported as using generative AI tools less often.
Privacy: One example that comes readily to mind is being cautious about sharing photos or personal information. The prevalence and consequences of deepfakes and misuse of photos of women and girls could — again quite understandably — make them more reluctant to use genAI tools on their phones.
Free Time: Another reason might simply be that women, who often work an unpaid “second shift” on responsibilities at home, may have less free time to use genAI tools (or other apps!) on their phones, or to learn & use genAI tools ‘on their own time’ for their jobs.
Cost: Women are generally paid less than men across the board, and worldwide to varying degrees. Not all genAI tools are free, and for those that offer free plans, they are often limited.
Appeal: I also wonder about the Meta ads Berkeley used to attract the Kenyans when evaluating their interest in learning use ChatGPT. Given the low percentage of women who responded (73.9% male), I’m curious about what care was taken to ensure the ads weren’t slanted towards stereotypically male interests. [It reminded me of older studies claiming that girls simply don’t like to play computer games as much as boys do; but when girls are offered games and characters that resonate with their interests and values, they play games just as much. And harassment in social game environments is recognized to impact gameplay and experiences.]
Workload and Productivity: It’s also worth noting that there’s a big disconnect between CEO expectations of AI benefits and people’s experiences with AI use, in terms of impact on productivity. Over half of people of all genders frequently experience burnout due to overload at work. Since it’s still common for women to be expected to handle ‘extra’ tasks at work, their workloads may already be even heavier. Those who decline to use AI may be trying to manage its impact on their workload!
"Despite 96% of C-suite leaders expressing high expectations that AI will enhance productivity, 77% of employees using AI say these tools have added to their workload, and nearly half (47%) of employees using AI report they do not know how to achieve the expected productivity gains."
I didn’t see any mentions of privacy, deepfakes, overall phone app usage, cost, or workloads in the study report.
My Experience
I’ve interviewed over 50 people worldwide since August 1 about how and why they do & don’t use AI in a variety of roles. This interview series wasn’t intended to be a rigorous research study; my goal is to find & amplify the voices of real people whose experiences with using AI - or not - deserve to be more widely heard. It’s obviously a much smaller sample, but it’s about 60% women, in a mix of technical and non-technical roles and across a wide range of ages & experience levels. And there’s obviously potential for some self-selection bias in who volunteers or agrees to be interviewed. But qualitatively, I’m not hearing gender-based hesitancy with genAI at the levels these studies are reporting.
Bottom Line: 3 Recommendations
If you’re interested in this topic, read the full study, and not just the article.
Regardless of your gender, I encourage you to look for ethically developed AI tools by companies who have shown that they care about addressing biases. The more we as consumers use our attention and money to support companies with the ethical behavior we want, the more likely we are to get ethical tools.
Stay tuned for the results of some data analysis on the first 50+ interviews from me and
in . In the meantime, check out my 2024 year-end retrospective on the first 36 AISW interviews.
References
This article is an updated version of this LinkedIn post, written after I read the full HBS study and some of the studies it evaluated. It also incorporates thoughts from a recent post on Substack Notes.
“Women Are Avoiding AI. Will Their Careers Suffer?”, featuring Rembrand M. Koning. By Michael Blanding / Harvard Business School, 2025-02-20. [full study: “Global Evidence on Gender Gap”]
“Upwork Study Finds Employee Workloads Rising Despite Increased C-Suite Investment in Artificial Intelligence”, 2024-07-23. [full study: “From Burnout to Balance: AI-Enhanced Work Models”]
Karen! What a great write up! I so appreciate you showing how you can be open to AI and still stand up for people's right to refuse it on ethical or professional grounds. And I totally agree, shaping tools and practices to respond meaningfully to these concerns will benefit us all. I can't wait for the results of your 50 person interview on this topic! Keep going! 👏👏👏
Very good article! I writing a take down of "Lean In" (not in the context of AI, but in how the concept confirms scarcity mindset) in a future substack post!