The Gender Gap in AI Is Real.

Here Is What To Do About It.

Last week, Reese Witherspoon posted a video that got over three million views. She had asked ten women in her book club how many of them use AI. Three said yes. Only one felt she actually knew what she was doing. She cited research showing that women use AI at a rate 25 percent lower than men, and that the jobs women disproportionately hold have higher exposure to automation than those held by men.
The video generated a significant backlash. Some pushed back on the environmental cost of AI systems. Some raised legitimate concerns about copyright and training data. Some questioned whether framing AI adoption as a feminist imperative was appropriate, given how much harm from these systems has landed on underrepresented communities. All of those criticisms deserve serious engagement.
But the data Reese cited is real. And as a woman who builds AI systems for a living, I want to talk about what it actually means.

The gap is not about ability. It is about access, confidence, and who the technology was designed to speak to in the first place.

What the Data Is Actually Saying

The 25 percent usage gap is striking on its own. But the more important number is the one sitting underneath it: the jobs most exposed to automation are disproportionately held by women. Administrative roles, customer service, scheduling, coordination, data entry, the kind of work that has historically been undervalued precisely because it looks like it does not require specialized skill. AI is coming for that work first and fastest.

That creates a specific kind of vulnerability. Women who are not building fluency with AI tools are not just missing a professional development opportunity. They are at greater risk of being displaced by the very systems they have not learned to use. And the businesses founded or led by women that are not integrating AI are running at a structural disadvantage against competitors who are.

That is the stakes version of the gap. The quieter version is something I see more often in the rooms I actually work in: women who are curious about AI, who know they should be engaging with it, and who do not know where to start in a way that does not feel overwhelming or beside the point.

Why the Gap Exists

The simplest explanation is also the truest one. Most AI tools were built by people who were not thinking about the full range of people who would need to use them. The early adopter culture around AI has been overwhelmingly male, overwhelmingly technical, and overwhelmingly oriented toward engineering and coding use cases. The language, the examples, the communities, the onboarding, all of it was designed for a user that looked a certain way.

That does not mean the tools do not work for everyone else. It means the entry point was not built with everyone else in mind. And for someone who is already stretched thin running a service business, managing a team, and trying to figure out which of the forty AI tools being released every week actually matters, a difficult entry point is often enough to make the whole thing feel like it is not for her.

There is also something subtler happening. The way AI has been marketed, as an optimization tool, a productivity multiplier, a way to do more faster, speaks to a very particular kind of ambition. It does not speak as readily to the kinds of value that many women-led businesses are actually built around: the quality of the relationship, the depth of the service, the feeling a client gets when they work with you. Those things matter enormously in AI strategy, but you would not know it from most of the content that exists about AI for business.

The most important thing AI can do for a service business is protect the human work, not replace it. That reframe changes who the technology is for.

What I Have Seen in Practice

I build AI systems for a living. My co-founder Ashley and I work primarily with small and mid-size service businesses, and a significant portion of our clients are women. What I see consistently is that once the entry point shifts, once we stop talking about AI as a productivity tool and start talking about it as a way to protect the work that only you can do, everything changes.

The woman who was not interested in AI for its own sake becomes deeply interested in an AI system that handles her follow-up sequences so she can spend that time on the client call that actually builds the relationship. The founder who felt like AI was for tech companies becomes genuinely excited when she sees what a well-built email nurture workflow can do for her conversion rate without changing how she shows up for her clients.

The gap is not about ability or interest. It is about framing. And the framing that most AI content leads with is the wrong one for most of the women I work with.

The Criticism Worth Taking Seriously

The backlash to Witherspoon's video raised three criticisms that I think are worth engaging with directly rather than dismissing.

The environmental argument is real. AI systems consume significant energy and water resources. The scale of that consumption is growing, and it is not evenly distributed in terms of who bears the environmental cost. For anyone building a business that cares about sustainability, this is a legitimate factor in how you evaluate which tools to use and how much. It does not mean the answer is to avoid AI entirely. It means energy and resource use should be part of the calculation.

The copyright and training data concerns are also legitimate. Much of what large language models know, they learned from work created by people who were not asked and were not compensated. That is an unresolved ethical and legal question, and it will not be resolved quickly. Being aware of it, using tools responsibly, and advocating for clearer standards is a more useful response than either ignoring it or letting it stop you from engaging with the technology at all.

The most substantive criticism is the one about harm. AI systems have demonstrated real bias against underrepresented communities, in hiring tools, in content moderation, in facial recognition, in credit and lending decisions. Women, and particularly women of color, have been disproportionately affected. That history matters. It should shape how we evaluate tools, how we advocate for accountability from the companies building them, and how we think about the use cases we choose to build.

None of that changes the reality of the adoption gap. If anything, it makes the case stronger. The people with the most at stake in how AI develops are precisely the people who are least represented in the rooms where those decisions are being made, including the rooms where AI tools are being learned and deployed. Stepping back from the technology does not protect you from it. It just removes your voice from the conversation.

The people with the most at stake in how AI develops are precisely the ones least represented in the rooms where it is being built. Stepping back does not protect you. It removes your voice.

What To Do

The solution to the gap is not a crash course in ChatGPT. It is something more deliberate and more personal than that. Here is how I think about it for the women and women-led businesses I work with.

Start with one specific problem, not the technology

Do not start by asking what AI tools you should be using. Start by asking where in your business you are spending time on work that does not require you specifically. That is the right entry point. The tool question comes after the problem question, not before it. If you start with the tool, you end up with a solution looking for a problem. If you start with the problem, you end up with something that actually changes your day.

Reframe what you are protecting

The most powerful shift I have seen in how women engage with AI is when the frame moves from "what can AI do for me" to "what does AI allow me to protect." The client relationship. The creative work. The judgment calls that require your specific experience and your specific understanding of your clients. AI can handle a remarkable amount of the work that surrounds those things. When you see it that way, the technology stops feeling like a threat to what you are good at and starts feeling like a tool in service of it.

Close the gap with a specific workflow, not a mindset shift

Learning AI in the abstract is not a strategy. Pick one workflow in your business that is repetitive, time-consuming, and does not require your direct involvement. Build one system around it. Learn what works and what does not. Then do it again. That is how fluency develops, not through a course or a summit or a commitment to staying current, but through the accumulated experience of building things that work and adjusting the things that do not.

Build in community

One of the clearest patterns in the research on women and technology adoption is that peer learning environments outperform individual learning environments significantly. Find two or three other women who are at a similar stage with AI and create a deliberate space to share what you are building, what is working, and what is not. The knowledge gap closes faster in community than in isolation. This is not a soft observation. It is a structural one.

Where I Land

I am not interested in telling women they need to learn AI because the machines are coming. That framing is both inaccurate and unhelpful. I am interested in making the case that the tools that exist right now, used with intention and applied to the right problems, can meaningfully change what it costs to run a service business, how much time you spend on work that drains you, and how much space you have for the work that actually matters.

That case does not require urgency or fear. It requires honesty about what the gap is, where it comes from, and what actually closes it. The gap is real. The tools are accessible. And the entry point, for anyone who has been waiting to find the right one, is a single workflow in a business you already know better than anyone else does.

That is where it starts.

Chantal Emmanuel is the founder of BAMPT, where she helps service businesses implement AI-powered operations using the A.G.E.N.T. Framework. She’s also CTO of LimeLoop and writes about automation, systems thinking, and building businesses that scale.