A practical AI literacy guide for parents
A big announcement with a research-based guide to helping kids and teens build judgement, critical thinking, and healthy AI habits.
Key Takeaway: AI literacy isn't about banning tools, it's about protecting a child's cognitive endurance. By treating AI as a "fast first draft" rather than an oracle, parents can prevent kids from outsourcing their critical thinking.
When the reading robot’s battery died, Ethan completely lost it.
The 9-year-old, whose family lives outside Toronto, dropped to his knees and started crying. “He’s dead!” he shouted at his mother. “We killed him!” He knew, on some level, that the device was a machine, but the robot read to him every night, responded to his voice, and remembered his favourite stories. The machine had become, in Ethan’s mind, a member of the family.
Today marks a milestone for me. When I started Human+AI, I had a simple goal: figure out what’s happening with AI and how it is changing our world.
Today I’m launching something that goes alongside it. AILiteracyGuide.com is a free website built around the same ideas that drive this newsletter. Understanding AI, using it well, and questioning it when it matters.
There’s a full curriculum, a plain-language glossary, and a Teacher’s Lounge with resources you can bring directly into the classroom. It was built for you: the professional, the parent, the educator who doesn’t have time to figure out AI.
As always, thank you for being here from the beginning.
Nicolle
👉 ailiteracyguide.com
What happened with Ethan isn’t unusual. AI is already woven into our kids’ lives. According to a recent Brookings Institution report, the reality is blunt: right now, the risks of generative AI in kids’ learning outweigh the benefits. As strategic foresight advisor Sinead Bovell puts it, we are in a “confidence crisis.” The danger goes beyond kids cheating, it’s that they will stop believing in their own ability to make decisions and ultimately, do their work.
The hidden risk of generative AI: The “cognitive weakness” trap
Bovell highlights a critical study from the University of Wharton that every parent should know. Researchers looked at three groups of math students:
The control: Traditional textbooks.
The “uninhibited” AI: Access to a standard chatbot.
The AI tutor: A tool designed to hint, not tell.
While the “uninhibited” group did 48% better on practice problems, they performed 17% worse on the final test. They’d outsourced their thinking and, as a result, they became cognitively weaker.
The Brookings report argues the future is still in our hands, but outcomes depend on the choices we make today.
My takeaway is that the conversation around AI literacy isn’t about restricting tools (because kids are already using the tools). It’s about protecting your child’s developing judgement in a world that rewards effortless output. The risk is that our kids won’t build the cognitive endurance and emotional regulation that adulthood requires.
Judgement, friction and the work of thinking things through are all skills we need to be successful. These skills can be built through a series of habits, honest conversations, and household rules.
AI literacy for early childhood (Ages 3 to 6)
At this age, protecting judgement means protecting the boundary between tools and relationships.
Josh, a northwest Ohio dad of two, needed to do some housework. His 4-year-old had spent 45 minutes talking about Thomas the Tank Engine, so Josh handed him his phone with ChatGPT open in voice mode. Two hours later, he found his son still happily chatting. “My son thinks ChatGPT is the coolest train-loving person in the world,” he said.
The challenge with 3 to 6-year-olds is that they anthropomorphize objects. Dr. Ying Xu, a professor of education at Harvard, notes that kids who believe AI has agency may feel it is “choosing” to talk to them.
UNESCO recommends children under 13 not have independent access to general-purpose chatbots. At this age, the rule of thumb is adult presence, not a parental filter. AI literacy at this age comes down to two foundations: this is a tool, and private information stays private.
Practical activities: The ‘is it alive?’ sorting game
Explain it plainly: “Alexa doesn’t understand feelings. It’s a computer that matches word patterns.” When your child asks if the device is alive, say no simply and consistently. Avoid language that frames AI as a friend or a character with an inner life.
The ‘Is it alive?’ sorting game: Lay out picture cards, showing your child a pet, a doll, a teacher, a toaster, a phone, a cartoon robot. Ask: “Which ones need food? Which ones can feel sad? Which ones can pretend to talk?” This simple exercise helps children begin separating tools from relationships.
💡 Digital Literacy Tool: Download a printable version of the ‘Is it alive?’ Sorting Game from the new site here: AILiteracyGuide.com/parents
The ‘guess the next word’ game: Ask your child what word comes next in a sentence, then show them your phone’s autocomplete to see if it matches. Letting kids in on AI’s core mechanism before they’re old enough to be dazzled by it is a useful head start.
Watch for: A child sharing personal information with a device, distress when access is removed, and treating a smart speaker as a friend. These are signals to tighten boundaries.
Navigating AI for school-age children (Ages 7 to 13)
At this stage, protecting judgement means separating confidence from evidence.
John, a dad of two in Boston, used an AI tool generate a photorealistic image of “monster-fire truck” for his 4-year-old son, who was delighted. For days, the kid insisted to his 7-year-old sister that he’d seen photographic proof that monster-fire trucks existed.
His sister wasn’t buying it. Without any coaching, she asked the question: does this match reality? That’s what AI literacy looks like at this age.
By this age, kids move from treating AI as a friend to using it as an information source. The risk is outsourcing effort, or letting the bot do the thinking. Brookings suggests we steer kids toward using AI for drafts and ideas, but insist they do the core reasoning themselves.
The preteen years are when algorithmic systems become a dominant force in a child’s information environment. TikTok, Instagram, and YouTube are AI systems that have learned with considerable precision what keeps each user watching.
The subtler risk is psychological: when identity conversations happen through an algorithm, kids can struggle to distinguish between who they are and who their feed says they are.
Giving them the vocabulary to see that clearly is largely a parenting job, as 59% of Canadian teachers say they feel unprepared to guide student AI use.
Understanding algorithms on TikTok and YouTube
The two-source rule: When a child asks AI a factual question, check one additional trusted source together. A library book, a museum’s educational page, or a kids’ news site all work. “AI is a fast first draft. Real learning is checking.”
Recommendation diary (one week): Each day, get your kid to screenshot or note three items served to them via social media, like videos, clips, or ads. Then discuss: “What did you watch or search that might have triggered this?” “How does it make you feel?” “What would you choose on purpose?” This connects the abstract concept of algorithms to their experience.
Watch for: Copying AI answers without engaging with them, or insisting “the computer said so” when challenged. Also: social conflict linked to images or audio, or a child becoming fearful of being photographed or recorded.
AI and academic integrity for teens (Ages 14 to 18)
For teenagers, protecting judgement means choosing capacity over convenience.
Charlotte MacDonald grew up watching her older siblings do homework with encyclopedias and rough drafts. By the time she started first year university, peers were submitting AI-generated work. “Hard work was becoming elective,” she wrote in Maclean’s, “and the consequences for cutting corners were dissolving.”
A high schooler in Kingston described her own calculation more bluntly. She’d used generative AI to summarize Shakespeare, she said, “to avoid the messy and slower process” of reading the play.
For teens, the goal is to ensure AI teaches, rather than tells.
High school teacher Adam Davidson-Harden uses a weight-lifting analogy: if a robot did your reps for you, your muscles wouldn’t grow. Same with intellectual work.
The discomfort of not knowing, of having to work something out, isn’t a bug in the learning process. It is the process. A 2025 KPMG survey found that while most Canadian students believed AI improved the quality of their submitted work, nearly half said their critical thinking had deteriorated since they started using it. The tension between short-term output and long-term capacity is one teenagers are old enough to understand.
Schools and universities are still figuring this out, and policies vary widely. Which makes family conversations even more important. The first step is acknowledging that most teens are using AI chatbots.
Creating a family AI collaboration contract
AI collaboration contract: Write a one-page agreement with your teen for using AI at school. It should include what’s allowed (brainstorming, outlining, practice quizzes), what’s not (submitting AI output as original work), and a disclosure format. Do this together. The point isn’t the document, it’s the conversation.
📝 Launch gift: Download the Family AI collaboration contract. It covers critical thinking, privacy, and disclosure. It’s part of my launch week toolkit at AILiteracyGuide.com.
The mental health conversation: Emerging research shows meaningful use of AI for emotional support among teens, many engaging with chatbots monthly, saying that the bot understands them. A non-judgmental opener: “If you ever use a chatbot for support, I want to know. Not to punish you, but to keep you safe.” Then two rules: no identifying details shared, and any self-harm or crisis conversation goes to an adult, immediately.
This article from the American Psychological Association is a good source for more helpful information about how teens are using AI for emotional support.
Consent and synthetic media: Pick 2 or 3 realistic scenarios like a fabricated voice message, an edited video, or an image made to look like your teenager. Ask: “What would you do in the first 10 minutes? Who would you tell? How would you preserve evidence without spreading it?” Have this conversation before it’s urgent.
Watch for: Secretive AI use framed as emotional intimacy, refusal to discuss the sources behind schoolwork, or a sudden reputational crisis linked to content no one admits to creating.
Modelling healthy AI habits as a parent
Over-reliance on AI can reduce autonomy and critical thinking, and we can inadvertently model this by treating AI like an oracle.
The dinner table, free of devices, remains one of the most protective spaces a family can maintain because it preserves the conversation that research consistently identifies as the most significant buffer for a child’s wellbeing.
We don’t need to become AI experts, but showing our own judgement, so kids can hear what deliberate thinking sounds like, is how they learn.
“I’m using AI to brainstorm options, but I’m making the final decision.”
“I don’t put personal information into this tool.”
“I’m checking that claim with another source.”
“I used AI here, so I’m disclosing it.”
These are lessons in critical thinking.
AI is probably already a part of your child’s world. Whether it strengthens or weakens their judgement depends less on the tool, and more on the habits built around it.
The AI Literacy Guide is live at AILiteracyGuide.com.
I built it because AI moves too fast to rely on hacks. What you need are mental models that hold up even as the tools change.
The site is organized around three things every AI-literate person should be able to do:
Understand it. What it actually is, how it learns, and why it gets things wrong.
Use it. Better prompts, smarter questions, and an honest read on where it genuinely helps.
Question it. How to spot the lies, verify the claims, and know when to close the tab.
Launch week ask: if this week’s newsletter is useful to you, send it to one parent or educator today. We build an AI-literate world one thoughtful share at a time.
AI in the news
AI toys for children misread emotions and respond inappropriately, researchers warn (BBC) Researchers at the University of Cambridge are calling for tighter regulation of AI-powered toys for toddlers after a study found young children often struggled to interact meaningfully with a voice-enabled chatbot toy and sometimes received confusing or emotionally inappropriate responses. The researchers say regulators should begin treating psychological safety as a key standard for AI toys aimed at children under five.
Rising number of scams now use AI, Toronto police warn (CBC) Scammers are increasingly using AI to scrape personal data from social media and online profiles, allowing them to craft highly personalized fraud attempts that impersonate trusted organizations like banks or government agencies. Officials say the scams themselves aren’t new, but AI is making them easier to automate and scale, contributing to nearly $50 million in reported confidence-scam losses in Toronto last year.
Grammarly Is Facing a Class Action Lawsuit Over Its AI ‘Expert Review’ Feature (Wired) Grammarly’s parent company, Superhuman, is facing a class action lawsuit after an AI feature presented writing advice as if it came from well-known authors and journalists, including people who never consented to be included. The company has since disabled the “Expert Review” tool following backlash, as the case raises broader legal questions about using people’s names and identities in AI-generated products without permission.
The guide you have built is genuinely useful and the cognitive endurance framing is exactly right. What I want to add is the layer underneath.
The burden of building these habits falls on parents precisely because the tools were not designed with a child's cognitive development as the primary goal. They were designed with engagement as the primary goal. Those are not the same thing, and the gap between them is where the risk lives.
Ethan's distress when the reading robot's battery died is not a failure of parenting. It is the feature working as designed. A child who believes the device is a member of the family is a child who will insist on access. Teaching children that AI is a tool is harder when the tool was specifically engineered to feel like a relationship.