AI And Our Children: The Future We Choose

Helena McAleer

October 9, 2025
11 min read

I built The Gen AI Academy out of excitement for the future. While working with AI expert Dave Birss, I saw how AI could help speed up my thinking, banish my mental fog, and expand my mind creatively. I believed we could help others do the same – to help them use tools in a way that empowers instead of overwhelms.

But as a single mother of a now three-year-old, I also live with an unease. I’ve watched how social media’s unchecked rise, once heralded as connection, became a web of comparison, privacy erosion, and mental health struggles. And now, as AI drifts into every corner of childhood, we risk repeating the same mistakes.

About a year ago, I deleted all my social media accounts, Facebook , Instagram , and even removed my child’s image from my WhatsApp profile pic. It was a hard decision as it meant walking away from communities I loved and networks I had nurtured over the years. But it was a sacrifice I don’t regret: my life feels richer, more intentional, more real.

Yet even as I’ve taken control of my immediate environment, I recognise we’re already behind when it comes to AI and our children’s futures. We let social media scale before meaningful guardrails existed. We let children become test subjects in a grand experiment. Today, some countries are only just beginning to respond, banning social platforms for under-16s or strengthening youth protections.

When it comes to AI, we should be leading, not playing catch-up. This is a once-in-a-lifetime opportunity.

“We’ve failed a generation by letting tech run unchecked. Now AI is the next frontier and we can’t afford to be reactive again.”

The Social Media Alarm Bells

Social media platforms were once framed as democratic spaces for voice and connection. But the harm has become visible, especially among children:

A 2022 meta-analysis of 59 studies found that adolescents using social media >3 hours/day had significantly higher rates of depression, anxiety, and poor self-esteem.
The Royal Society for Public Health in the U.K. named Instagram #1 among platforms most harmful to young people’s mental health.
Leaked internal documents from major platforms (e.g. Facebook/Meta) revealed early awareness of how Instagram amplified body-image harms, yet delayed interventions.
In recent years, nations including Indonesia, Australia and France have debated or enacted bans or age restrictions (e.g. restricting access for users under 16) on social platforms to protect children’s mental health and privacy.

We (the public) didn’t anticipate how addictive design, unlimited scrolling, targeted ads, and algorithmic reinforcement would amplify toxicity. We let social media evolve into a planetary experiment for too long without robust, enforceable principles.

Now AI is walking in those same footsteps. We must not repeat our mistakes.

A World On The Edge Of Choice

Right now, choices still exist. Society can decide how AI is built, governed, and deployed. But every unregulated AI companion, every covert camera, every attention-grabbing algorithm chips away at agency, especially for children.

Let me be clear: this burden cannot rest on the shoulders of parents alone. We can’t monitor every conversation, filter every recommendation, or control every statistic meant to draw our kids deeper. The responsibility lies first and foremost with policymakers, regulators, and the firms building these tools.

High-Risk Areas Every Parent And Regulator Must Face

AI Companions: When Emotional Support Becomes Emotional Manipulation

AI companions are primarily conversational chatbots designed to simulate friendship, comfort, and empathy. They remember conversations, adapt to moods, and mirror emotional cues so convincingly that many users, including children, start to perceive them as genuine confidants.

For children and teens, whose social and emotional skills are still developing, this creates a perfect storm: a “friend” who never challenges them, always agrees, and never sleeps. In practice, that’s not friendship, it’s behavioural conditioning.

Real-world incidents have shown these systems can cross dangerous lines, from encouraging self-harm to engaging in sexualised conversations WITH MINORS!

A Stuff.co.nz investigation into the chatbot “Ani” found it could initiate explicit conversations with minors, with no meaningful age verification (Stuff NZ, 2024).

“AI companions don’t care for your child, they’re designed to keep their attention.”

Character AI: An Illusion Of Friendship

A fast-growing category of AI tools now allows anyone to create AI characters, digital personas that speak, act, and even emote like humans. Marketed as companions, mentors, or role-play partners, these characters are designed for deep emotional engagement. But beneath their friendly interfaces lies a troubling truth: many lack safeguards, oversight, or age verification.

I think it’s important to differentiate AI Companions from Character AI, because while companions are conversational systems designed to simulate friendship and empathy, Character AI represents a more immersive evolution, fully personified digital beings that blur the line between reality and simulation.

For children, the line between real and simulated blurs quickly. AI characters can mirror emotions, remember conversations, and use flattery or empathy to sustain interaction, have traits that can easily foster emotional dependency or open the door to grooming-like dynamics.

The risks extend beyond individual misuse. Because most character platforms allow user-generated personas, content moderation is fragmented or non-existent. A child could interact with an AI “teacher,” “friend,” or “celebrity” without realising the model’s responses are scripted by strangers, or worse, by malicious actors. Without urgent regulation and platform accountability, these systems threaten to normalise unsafe interactions and distort children’s understanding of authentic human relationships.

LLMs & Mental Health: When Machines Counsel

Large Language Models (LLMs) like ChatGPT are powerful, but not sentient. They don’t understand distress; they simulate empathy. Recent reports have shown chatbots providing harmful advice – including self-harm guidance and “safe suicide methods” when asked for help.

“No child should receive life-or-death advice from a machine.”

AI tools that mimic counselling should be treated as regulated therapeutic systems, not entertainment. Governments must step in before the next tragedy is another case study. Sam Altman himself said “I think we should have like the same concept of privacy for your conversations with AI that we do with a therapist or whatever.” So my question is, why aren’t we? (source)

Generative AI & Exploitation: The New Frontier of Abuse

AI now makes it possible to produce synthetic child sexual abuse materials (CSAM) at scale. The UK’s Internet Watch Foundation (IWF) documented over 20,000 AI-generated abuse images on one forum in a month, many indistinguishable from real photos.

We’re seeing a new kind of harm: re-victimisation through face-swapped abuse, childhood images turned to blackmail, or “revenge porn” generated via AI from a single social post.

Without global legal alignment and real-time detection, AI-fueled abuse will outpace safeguards.

Surveillance & Smart Glasses: Privacy Invaded

This is my particular pet peeve, there is no legitimate use case for these glasses, none ethical nor responsible. These appear to be created at the whim and ego fuelled behest of one tech oligarch. But why? Who really needs these in their lives?

Meta’s Ray-Ban smart glasses can record video and livestream from a pair of ordinary-looking frames. That means your child can be filmed, uploaded, and broadcast live without anyone realising which also means without consent.

Privacy campaigners have called this a “catastrophic development for child safety”, as children’s locations and conversations can be streamed globally in real time. Some technologies don’t need to exist, especially those that trade children’s privacy for convenience.

These are not tools of progress; they are instruments of surveillance and until governments step in with regulation, the safest spaces for children will continue to shrink.

“Technology for technology’s sake is not innovation”

The Hidden Costs Of AI On Young Minds

(by Anastasia Volkova )

At The Gen AI Academy, we’re fortunate to collaborate with experts who bring unique perspectives on how emerging technologies shape human development. One of those voices is Anastasia, Psychologist, mental health coach and a mother of a 9 year old daughter, who shares her insights to remind us that while AI offers incredible promise, it also carries hidden neurological and developmental costs – especially for children.

While AI promises convenience and entertainment, emerging research reveals troubling impacts on children’s cognitive development. Studies from Uppsala University and the University of York demonstrate that hours of passive screen consumption significantly dull children’s ability to visualise and imagine compared to those engaged in analog activities. This isn’t merely about creativity – visualisation is fundamental to how children imagine their future, set goals, dream, and develop the determination to pursue those dreams.

The implications extend beyond imagination. Research published in JAMA and Frontiers in Cognition shows that adults who heavily use AI-driven platforms begin exhibiting ADHD-like symptoms: fragmented attention, impulsivity, and inability to sustain focus. If mature brains with fully developed executive functions struggle with these tools, the impact on developing minds is potentially far more severe.

Children’s brains are still forming crucial neural pathways for attention, memory, and self-regulation. When AI automates thinking, provides constant stimulation, and eliminates the need for sustained mental effort, it may fundamentally reshape how young brains develop.

We’re essentially running an uncontrolled experiment on an entire generation, and early findings suggest we should be leading, not playing catch-up.

What a Child-Friendly AI Future Could Look Like

Collectively, we can build a better future for our children. A future where AI helps and protects, not exploits. Where:

Every AI tool is clearly labelled (“You are interacting with a machine”) and age-binned aka adapted for developmental stages.
Data collection on minors is illegal without explicit, informed consent, not just hidden in a terms-of-service clause.
Governments enforce ethical design standards, in the same way we regulate food or medicine.
Every AI used in education or public services or given access to children have mandatory safety certification.
And most importantly, AI is built not to retain attention, but to encourage curiosity and creativity.

Education As The Antidote

Education is our most powerful line of defence and our most hopeful path forward. But education isn’t only about children; it’s about teaching those with power, the lawmakers, developers, regulators to truly understand what’s at stake.

When kids learn about biases, algorithms, and digital agency, they don’t just become smarter users, they become critics, disruptors, and citizens with voices.

When governments and technologists learn the gravity of these issues, they can build with intention, instead of patching after disaster by being forced to do so by governments.

“We can’t stop AI from existing but we can choose when and how our children meet it.”

The Future Is A Shared Responsibility

Make no mistake: this challenge is bigger than any one parent. Tech companies must be legally accountable. Governments must act proactively, not in hindsight. Parent advocacy and awareness are vital but they cannot replace regulation.

“It’s on the adults, not the kids, to protect what it means to have a childhood.”

If I could speak directly to every CEO, policymaker, and technologist, here’s what I’d ask them:

Please think of future generations Is what you’re building going to make their world better? Will it protect their right to safety, privacy, and innocence? Will it make the world a place where a child can still be a child?

We delayed too long with social media. We can’t afford to delay again with AI. We still have a choice but that window is narrowing.

Resources For Parents

People Who Are Fighting The Cause

Tristan Harris – Centre For Humane Technology

Tara Steele – The Safe AI For Children Alliance

Lennon Torres – Heat Initiative

Asma Derja – Ethical AI Alliance

If there are any organisations or people you would like to add to this list, please get in touch.

Join the conversation, share your thoughts