Women, AI, and the Room We Deserve to Be In

Recently, Reese Witherspoon posted an Instagram reel. She'd noticed that at a recent book club gathering, only 3 in 10 women were using AI, and even fewer felt confident they were using it well. Her take: women need to get more familiar with this technology.

The backlash was swift. People called it disappointing. Commenters told her to start her AI education with data centres and energy consumption. The discourse, predictably, collapsed into two camps: AI boosters and AI sceptics, with women expected to pick a side.

Here's what I think: both camps are missing the point.

The concerns are real. That's exactly why we need to be in this.

The criticisms of AI aren't wrong. There are genuine, serious issues, such as copyright violations, algorithmic bias, environmental cost, labour displacement. These aren't fringe concerns. They're documented, researched, and ongoing.

But here's the thing that doesn't get said enough: women's jobs are three times more likely to be automated by AI, while women currently use and engage with AI tools 25% less than men. If both of those things are true at the same time, disengagement isn't a principled stance. It's ceding the ground to people who are not thinking about our interests.

The argument for getting informed because AI has problems is actually stronger than the argument for getting informed because it's exciting. We don't get to shape what we don't understand.

Women built this field. We just keep being written out of it.

There's a persistent myth that technology, like computing, software, AI, has always been a male domain. It's a myth worth dismantling, because the history is genuinely different.

Ada Lovelace published what is considered the world's first algorithm in 1843, envisioning that machines could do far more than arithmetic decades before a working computer existed. The women of ENIAC programmed the first general-purpose electronic digital computer, not as a footnote, but as the inventors of how programming itself worked. Grace Hopper, known as the "Queen of Software," wrote the first computer manual, developed early programming languages, and is credited with coining the term "debugging" after tracing a system fault to an actual moth lodged in the hardware.

The "women aren't tech people" story is historically recent and factually wrong. Computing used to be women's work until it became prestigious and well-paid, at which point the history got quietly rewritten. Recognising that isn't just an act of reclamation. It's accurate.

The women shaping AI right now

Far from being absent from AI, women are among its most consequential architects, both building it and holding it accountable.

Joy Buolamwini and Timnit Gebru didn't just critique facial recognition — they proved it was broken. Their Gender Shades research revealed error rates of under 1% for light-skinned men compared to nearly 35% for dark-skinned women in widely-used commercial systems. That research forced IBM and Microsoft to act. They didn't wait to be invited to the table. They built the evidence that made the table impossible to ignore.

Timnit Gebru went on to found the Distributed AI Research Institute (DAIR), an independent organisation dedicated to ethical AI development free from corporate and government pressure. When she was pushed out of Google for raising bias concerns, she didn't go quietly.

Daniela Amodei co-founded Anthropic after serving as VP of Safety and Policy at OpenAI. Her path went from English literature to congressional campaigns to Stripe to the frontier of AI development, and she's a living argument that you don't need a computer science degree to help determine what this technology becomes.

Mira Murati led the development of ChatGPT and DALL-E at OpenAI before launching Thinking Machines Lab in 2025, raising a record-breaking $2 billion seed round. She is building AI with interpretability and safety at its core.

Fei-Fei Li, often called the Godmother of AI, created ImageNet, the dataset that underpins most of modern computer vision and deep learning. She co-directs Stanford's Human-Centred AI Institute and founded AI4ALL, a nonprofit working to make AI education more inclusive.

Sinead Bovell is a futurist and founder of WAYE, making conversations about AI and the future of work accessible, especially for younger audiences who are inheriting this world. Her podcast I've Got Questions is one of the best places to start if you want honest, human takes on where all of this is going.

Nadia Lee is an Australian AI ethics founder whose work at ThatsMyFace was born from a simple, alarming realisation: we never consented to having our faces used as training data. Her work sits at the intersection of digital rights, ethics, and practical accountability, and she's doing it from here.

These women are building the infrastructure, writing the policy, and doing the research that determines what AI does and doesn't do to the rest of us. The question isn't whether women belong in AI. It's whether the rest of us are paying attention.

What's actually in it for you

I want to be clear about something: I'm not asking women to carry the burden of fixing AI's problems. That framing, women need to be in AI so they can be the ethical voice in the room, still puts the weight of a systemic problem on individual women. That's not the ask.

The ask is simpler and more personal. AI can work for you. Here's what I mean.

Understanding your own health. Women are historically under-researched and under-believed in medical settings. Conditions like endometriosis take an average of seven to ten years to diagnose. Walking into an appointment having actually understood your test results, knowing what questions to ask, knowing what a normal range looks like — that's a power shift. AI as a translator between you and a medical system not designed for you is genuinely useful.

Financial literacy, not financial advice. Women save more than men and invest less. That gap has real compounding consequences over a lifetime. Understanding what an index fund actually does, what compounding means in practice, what the difference is between a tax offset and a deduction — these are things many of us were never taught. AI is a patient, non-judgemental explainer for concepts you were somehow supposed to already know.

Advocating for yourself without the emotional overhead. This is the one I keep coming back to. Think about how many times you've softened an email before sending it. Added an apology you didn't mean. Started a question in a meeting with "this might be silly, but..." Spent twenty minutes rewording a message to a manager so it didn't come across as demanding.

AI has no feelings to manage. No relationship to protect. No face to read.

You can type: "Help me write an email asking for a 20% pay rise. Make it direct and confident." No preamble. No softening. No worrying about how you're coming across. And then you can push: "more assertive." "Too much, pull it back." You can practise being direct in a space with zero social consequences, and then take that directness into the room where it counts.

The emotional labour of managing how you're perceived while simultaneously trying to advocate for yourself is exhausting. It is not equally distributed. And it is one thing, not the only thing, but one real thing, that AI genuinely doesn't ask of you.

None of this means trusting AI blindly. It gets things wrong. It can reflect the same biases we're trying to push back against. A medical explanation still needs a doctor. A financial concept still needs your own judgment. What AI can do is save you hours of searching and give you a starting point, but the critical thinking, the final call, the knowing-when-to-push-back, that stays with you. Which is, in the end, another reason to learn how it works. The better you understand it, the better you can use it, question it, and know when to put it down.

So here's where I land

Reese Witherspoon wasn't wrong that women should learn more about AI. The backlash wasn't wrong that AI has serious problems. But the choice between cheerleader and sceptic is a false one, and it's not a choice we can afford to keep making.

Women have been in this field since the beginning. Women are leading it right now! Women are building it, auditing it, holding it accountable, making it accessible. The technology has real problems and real possibilities, often at the same time.

You don't need to become an engineer. You don't need to love it. You just need to know enough to ask good questions, use it on your own terms, and — if you have the platform and the appetite — be part of deciding what it becomes.

That room is being built right now. We should be in it.

Want to keep going? Start with Sinead Bovell's podcast I've Got Questions, Joy Buolamwini's TED Talk on algorithmic bias, or just open a tab and ask something you've always wanted to understand.