Back in September 2025, Nicole O’Connor spoke at WomenEd NI's 7th Annual Autumn Unconference about a 'Feminist Approach to AI' and promised a follow-up series collaborating with local, disruptive women to bring this framework to life.

Virginia Méndez and Nicole kicked off Part 1 with Be a Critical Consumer.

Now, it's time to ask some hard questions of ourselves - and the algorithmic machine.

If Part 1 was the ‘how’, Part 2 is the ‘who’.

Who gets to be a part of this chapter in humanity’s story? Who is leading, creating, or educating? And crucially, who is being exploited, silenced, or forgotten?

Whose Story Matters?

Representation is everything. That’s something Orla McKeating, the co-author of this blog, champions daily. So here's our discomfort with AI: not everyone is represented in the AI-powered tools we use on a daily basis. It is hugely biased and we, the humans, have to add friction into our AI interactions to make sure we’re mitigating this.

When Orla’s son was small, she looked for books with characters and heroes who looked like him, and families who looked like theirs. As a single white mother of a Black child, she didn't find many. That gap is why Still I Rise Voices exists. Now, when she asks an AI image generator for "a doctor", "a CEO" or "a family", it hands back the same narrow picture. It's the same gap, only faster and at scale.

Whose Data Matters?

We’re a bit fed up hearing tired tropes like:

“Tech is neutral!”

“AI isn’t bad, people are!”

The internet isn't neutral, and nor are humans. So therefore AI or SI (Super Intelligence) can't be neutral either. For starters, the Global South has vastly different access to connectivity than most of us reading this. We know AI learns from what has already been published, and that has always pushed some voices to the margins. Most of it comes from the English-speaking, Western internet.

English makes up roughly 20-50% of internet content (sources and research vary widely on this - which in itself is indicative of bias and misconceptions around minoritised languages and multilingualism), yet only around 16% of the world speaks it as a first language. Whole communities, cultures and knowledge systems are written out of the training data.

Joy Buolamwini's Gender Shades research found that facial recognition systems failed darker-skinned women far more often than lighter-skinned men. Narrow data doesn't just reflect bias. It amplifies it and grows over time. AI-generated content becomes training data, so missing stories get harder and harder to find. For children who rarely see themselves reflected, that isn't a glitch. It's often their identity being edited out.

While regulations, legislation and systemic change are all needed to sufficiently account for these biases, individual intentionality is the immediate leverage point we can take action on.

Disrupting the Algorithm

So how do we use AI without losing the human soul in the story?

We believe you have to be intentional with what you use, learn, watch and share. It takes effort to resist AI algorithms being the only one behind the wheel controlling and directing your attention.

That is by design. Once you see it, you can’t unsee it. So we recommend you tackle this on two fronts: Personal Practice and Community Campaigns. Empower yourself to disrupt their agenda by adopting some personal digital habits and then look a little closer, with intention, at what is going on around you.

Personal Practice:

1) The Algorithm Audit: Start with a quick audit. Scroll one of your usual social media feeds for 5 minutes. Screenshot every video or post as you go - don't stop to think, just scroll like normal, but capture each post you view.

Whose faces, accents, identities and stories show up? Who is missing?

If you can, borrow someone else’s phone - ideally someone of a different age, gender or race to you - and do the same thing as you scroll their feed for 5 minutes.

Compare and contrast. What differs between the algorithms? What do you notice?

Now act with intention. Follow creators from communities different to your own. This helps expand the echo chamber your algorithm perfectly curates for you.

2) The Shadowban Buster: Keep a running note on your phone of creators, thought leaders and disruptors from backgrounds unlike your own. When you catch yourself doom-scrolling, redirect yourself to their pages instead. It's a worthwhile habit to have as you’ll soon notice that disruptive voices get shadow-banned or abruptly removed whilst we’re kept well distracted and may not notice.

Nicole once lost Catharina Doria from Instagram entirely - found her on LinkedIn and she confirmed Meta had deactivated her account. She's back now, but it was a wake-up call to stay alert to voices that are being censored.

3) Learn with WomenEd: One of the most empowering things you can do is to upskill yourself in AI and automation. Recently some WomenEd-ers joined CVP Group for panel discussions on Women & AI - linked below - where the topic of ‘Who Writes the Story?’ was explored in the lead up to the launch of the Level 4 Artificial intelligence (AI) and Automation Practitioner - a curriculum designed by an all-female team of diverse educators and WomenEd-ers. Watch the videos and consider applying for a funded place on the Level 4 course. Cohort 1 begins this November.

Community Campaigns:

1) The Family Check-In: At the Festival of Education, Matt Pinkett, recommended a 20 minute weekly routine of co-viewing social media with your child.

Scroll with your teen (because remember, currently, under 13s should not have a social media account - however perhaps your younger child is using apps like YouTube Kids). Ask them:

Why do you think you get shown this?

Why would that brand target you and not me? 

Do you think your sister/uncle/teacher gets this content on their feed?

What data does this platform likely have about you? 

Starting like this will help children curate social media feeds that are healthier and safer for them - just stay alert for those with multiple accounts on various platforms - they all need to be co-viewed.

2. Active Campaigner: There are many amazing humans and activists campaigning for change. Seek them out. Here’s a couple to get you started:

Cindy Gallop is a huge advocate for algorithmic equality on LinkedIn where male accounts are platformed more than female accounts.

Joy Buolamwini's ‘Algorithmic Justice League’ campaigns for equitable and accountable AI - something we can all get behind!

3. Demand Evidence-Informed EdTech: Moving from individual habits to institutional accountability requires reliable frameworks. Organisations like the Chartered College of Teaching are driving this through the EdTech Evidence Board Project, which evaluates EdTech products against rigorous, evidence-based criteria. Pushing for this level of critical evaluation helps ensure that the tools we bring into schools support rather than undermine inclusive practice.

The Smartest Glasses

Again, we’re not asking you to stop using AI-powered tools, but to approach them with intention and discernment. When you use AI, name the diversity you want to see and question the defaults. Better still, put the tools in the hands of storytellers from those communities.

While schools cannot shoulder all of society's challenges, recognising the equity gap in how students and the wider school community learn to critically assess AI is where meaningful whole-school policy and practice begins. Once you do that, your natural next step is to start ‘Championing Co-Creation’ (Part 3) and writing a different story for the community you serve. The soul of a story comes from the person telling it. Our job is to make sure the machine helps us along the way but never gets the final word.

If you haven’t noticed how algorithms influence your world view and obscure intersectionality, activate your child-like curiosity and start questioning everything.

Curiosity is resistance.

Because at the end of the day, your real smart glasses aren't the ones on the market.

They're the ones with an intersectional lens.

Orla McKeating | @Orla_McKeating

Nicole O’Connor | @DigiKnowNicole