The Algorithmic Ceiling: What Happens When Women Aren’t in the AI Room

The glass ceiling didn't disappear. It became code.

"It always seems impossible until it's done." — Nelson Mandela

I wasn't looking for a research project.

I was in Mumbai, between meetings, and I started asking women a simple question: How are you navigating AI?

The first woman I asked worked at the courthouse. She laughed before she answered.

"I travel four hours every day. Then I take care of my family. When exactly am I supposed to learn this?"

Think she might be an exception? She wasn't.

What I heard — from Mumbai to Arusha

Over weeks of conversations with women leaders — bankers, media executives, startup founders, corporate professionals — I kept hearing variations of the same story:

"I'm not using AI because I think I'm bad at prompts. With bad prompts, I won't get good answers."

"I work for a startup and they don't use it, so I don't have immediate need at work."

"I use the free version. I don't know where the data goes."

One woman, a CHRO at a startup bank, told me she wasn't using AI because her company hadn't adopted it yet. Another, a media executive, was using it for content generation — but when I asked about governance, she paused. "We don't have anything in place."

Then I spoke with women in Tanzania. Different continent. Different barriers.

In Arusha — sometimes called the "Geneva of Africa" for its concentration of international institutions — AI still isn't widely adopted. Infrastructure is unreliable; our video call kept cutting out. And when I asked questions, the women were hesitant to speak openly in a group setting. One asked if we could talk one-on-one instead.

She wanted to learn. She just didn't want to look uninformed in front of others.

That's when I realized: the barriers aren't just time, confidence, or access. Sometimes it's the safety to admit what you don't know and are still learning. And sometimes it's whether your internet connection can hold a conversation at all.

If the governance gap exists in Arusha — a hub for international policy — proximity to power doesn't equal access to protection.

The Algorithmic Ceiling

We spent decades breaking through the glass ceiling. Now there's a new one forming.

I call it the algorithmic ceiling.

Here's how it works:

AI learns from data. When that data reflects a world where women weren't fully represented — in leadership, in decision-making, in the datasets themselves — then the algorithms quietly inherit that imbalance.

The algorithms aren't biased because someone intended harm. They're biased because they learned from history. And history left a data trail: fewer women in leadership roles, fewer women in technical fields, fewer women's voices in the rooms where decisions were recorded. The algorithms didn't invent the imbalance. They inherited it.

Unlike the glass ceiling, which we could see and name and fight, the algorithmic ceiling is invisible. It's baked into code. It scales at machine speed.

Bias doesn't just persist. It multiplies.

The Glass Ceiling Becomes Code

And here's what makes it worse: when women don't understand how AI works — where their data goes, how decisions get made, what biases might be embedded — then they can't guide teams, make appropriate decisions, or protect themselves. They become subject to systems they had no part in shaping.

The Governance Gap No One is Talking About

Right now, AI governance conversations happen in boardrooms. At summits. In policy papers most people will never read.

Don't get me wrong — that's important work. But it's not reaching:

• The woman commuting four hours a day
• The startup employee waiting for her company to "figure it out"
• The corporate executive whose company has AI tools but no AI guidance
• The leader who thinks she's "bad at prompts" and has stopped trying
• The professional in a region where AI hasn't arrived yet but will
• The community where reliable internet is still a barrier

When I explained to the courthouse worker why I was asking these questions — that AI is data, that systems built without women's input will reflect that absence — she paused.

"No one has ever explained it to me that way."

That's the gap that needs attention. Not just access to tools. Access to understanding.

Governance Without Permission

For too long, the message has been: "We need more women in AI." More women in tech. More women on ethics boards. More women shaping policy.

That's true — however, it's not enough, and it's not fast enough.

The woman commuting four hours a day isn't going to join an AI ethics committee. The startup founder isn't going to wait for her company to develop a governance framework. The corporate executive isn't going to read the EU AI Act before her next meeting. The woman in Arusha isn't going to wait for perfect infrastructure.

They need something else: the understanding to govern their own AI use, right where they are.

Not join a committee. Not become a technologist. Just know enough to ask the right questions before clicking "accept."

And sometimes, they need someone they trust to ask those questions with — one-on-one, not in a room full of strangers.

The Question I Can't Stop Asking

The AI governance conversation is well underway. Regulations are forming. Policies are being written. Algorithms are being trained right now, as you read this.

But here's the question I can't stop asking: Who's actually being reached?

Not who's writing the policies. Who's being protected by them?

The woman in Mumbai won't read the EU AI Act before her next meeting. The woman in Arusha might not have stable internet to join a governance webinar. They're not in the room — however, the decisions made there will shape their futures anyway.

So I stopped asking how to get more women into AI governance. I started asking: How do we bring AI governance to them?

That question became LIGHTHOUSE.

Building it Differently

I've spent 20+ years leading enterprise transformations. Now I'm focused on a different question: how do we bring AI governance to women leaders who don't have ethics boards, compliance teams, or hours to spare?

LIGHTHOUSE is designed for women leaders everywhere — corporate executives, entrepreneurs, small business owners, professionals navigating AI without a playbook.

It's also designed to travel through trust, not mandates. One woman learns. She guides another. Governance spreads through relationships, one conversation at a time. That's how it reaches women who won't speak up in a group. That's how it reaches communities where infrastructure is still catching up.

Every time a woman uses AI with intention — protecting her data, asking the right questions, making informed decisions — then she becomes visible to the system. Her input trains the algorithm. Her voice enters the data.

Every time she doesn't, the ceiling gets harder.

This isn't about just protection. It's about presence.

If we don't bring governance to them in a form they can actually use — then the algorithmic ceiling locks into place. And this time, we won't be able to see what's holding them back.

Come with me

If this resonates — and if you're navigating AI without guidance, or you want to help other women do the same — I'm building something for you.

I'll be sharing more soon: what I'm learning from the field, practical guidance you can use now, and how to become a Lighthouse Keeper yourself.

Follow me here or connect on LinkedIn to stay in the conversation.

The woman who told me she's "bad at prompts" doesn't need a summit. She needs someone in her corner.

Maybe that's you.

✦ Human Wisdom. AI Research. Responsibility Reinforced.