Back for another week of trying to keep up with what AI is doing around the kids in your life. This week, more than 200 groups told YouTube to deal with AI junk aimed at kids, Oklahoma lawmakers pushed new limits on certain AI chatbots for minors, LEGO made the case that kids should learn how AI works and not just use it, and Brookings reminded us that children are meeting AI earlier than many adults realize. That was only some of the week’s headlines. So grab your coffee, preferably made by a human, and let’s get into it.

In case you missed last week’s episodes, here is something to listen to on your way to work or school before we jump into the news:

Also, a quick thank you. A lot of new readers are here because some of you have been sharing this newsletter, and that really means a lot to me. If it has been helpful, you can share it with a friend or coworker and earn a few screen-free or physical goodies through the referral program below. Now, on to this week’s news.

More than 200 child advocacy groups and experts signed an open letter to YouTube's CEO on April 1st, and no, it was not an April Fools’ joke. They want YouTube Kids to ban AI-generated “slop.” The letter was organized by the children’s advocacy group Fairplay and signed by major organizations, including the American Federation of Teachers, the American Counseling Association, and The Anxious Generation author Jonathan Haidt. The groups say cheap AI-made videos are flooding the platform, including creepy cartoon animal clips, fake educational videos with mixed-up facts, and repetitive content designed to keep kids watching. Advocates say YouTube is making money from this content while young children, especially those too young to read AI labels, are left unprotected. YouTube has said managing AI slop is a priority for 2026, but critics argue the current labeling system is not enough.

What parents and teachers need to know: If your kid watches YouTube Kids, keep an eye on it. The platform is safer than regular YouTube, but odd AI-made videos can still slip through. A good habit is to watch a few videos with your child each week, because you will usually catch weird content faster than the algorithm. Watch for jerky character movements, audio that does not match, repeated actions, or “educational” videos that sound confident but make no sense. If something feels off, report it in the app.

U.S. Congressman Josh Gottheimer (NJ-5) introduced the bipartisan Parents Decide Act, legislation that would require operating system developers like Apple and Google to verify a user's age at device setup, not app by app, but from the moment a new phone or tablet is turned on for the first time. Under the bill, parents would set content controls upfront, and those settings would flow securely to every app and AI platform on the device. That means a chatbot would actually know it's talking to a child. Right now, it doesn't. The bill would also prevent kids from bypassing age requirements with a fake birthday, which is currently all it takes to access platforms they're not supposed to be on. Gottheimer pointed to cases involving teens harmed by AI chatbots and social media algorithms as the driving force behind the legislation.

What parents and teachers need to know: The big idea here is fixing the problem at the device level instead of playing whack-a-mole with individual apps. If it passes, parents would set the rules once, and the whole ecosystem would have to follow them: social media, AI platforms, all of it. In the meantime, the current system still relies on your child typing in an honest birthday. Spoiler: they probably will not. I know I had an AOL Instant Messenger screen name when I was in 5th grade, and I doubt my parents knew. I know, I know. So talk to your kids now about what they are accessing and why the guardrails matter, because the law has not caught up yet.

-Dr. Tiffany Petricini, Professor and Co-Chair of Penn State’s AI Committee (Quote from AI for Kids Podcast)

New polling from the Oklahoma Institute for Child Advocacy found that, in a survey of 603 likely Oklahoma primary voters, 93% were concerned about AI-generated explicit images involving minors, with 81% saying they were very concerned. Lawmakers also introduced SB 1521, which would prohibit certain human-like AI chatbots from being made available to minors, require age certification measures, and allow penalties of up to $100,000 per violation. It is part of a broader push to hold companies accountable for AI tools that could put kids at risk.

What parents and teachers need to know: We joke about kids knowing more about tech than we do, but the dark side, deepfakes, creepy chatbots, stuff we didn’t even dream of, is moving faster than expected. The poll shows parents across the aisle are done waiting for someone else to fix it. Talk to your kids about what they’re seeing online, set the house rules early, and keep an eye on what your school or after-school programs allow.

LEGO Education’s VP of Product Experience, Andrew Sliwinski, argues that schools are moving too fast to use AI on children without doing enough to teach children how AI actually works. He says the rush to bring AI tools into classrooms, from adaptive tutors to chatbot-based instruction, is only half the job. The other half is building real AI literacy by teaching students concepts like data, probability, algorithmic bias, and computational thinking. He points to a LEGO Education survey showing that nearly half of computer science teachers do not feel confident teaching AI topics even after training, and makes the case that these skills should be treated as core literacy, just like reading and math. His bottom line is simple: if kids understand how AI works, they can help shape it. If they do not, they will only learn to use it.

What parents and teachers need to know: There’s a line in this piece that stuck with me: “In our obsession with what computers can do, we have lost track of what children are capable of.” Most of the conversation about AI in education is about reducing risk. This piece asks are we teaching kids how to think about AI, or just giving them tools and hoping it works out? The next time your child uses an AI tool, ask, “How do you think it came up with that answer?” A child who stops to question the answer is in a different place than one who just accepts it, and that habit can start long before computer science class.

-Dr. Sarah Zipf, Researcher (Quote from AI for Kids Podcast)

The Brookings Institution released a new report titled “Generation AI Starts Early.” The report documents how children encounter AI long before they start school. From AI-enabled baby monitors and smart plush toys to voice assistants like Siri and Alexa, children are meeting AI early. About six in ten parents of children ages 2 to 8 report their child interacts with a voice assistant at least once per day. YouTube's daily use among children under 2 has climbed from 24% to 35% over the past five years. The report also highlights a key developmental concern: young children are most likely to assign human-like thoughts and feelings to technology. In plain terms, toddlers may genuinely believe the AI understands them or cares about them, which creates an early foundation worth paying attention to.

What parents and teachers need to know: What gets me about this report is how normal it all feels. Of course a two-year-old is talking to Alexa. Of course a toddler thinks Siri is a person. I mean, imaginary friends are a thing. Listen to how your child talks to AI. Do they say thank you? Do they get frustrated when it does not respond the way they expect? You can keep it simple and say, “That is a computer. It is helpful, but it does not have feelings.” This helps them understand what they are interacting with and respond to it in a healthier way.

A survey of 9,000 public school teachers in England found that many think students are losing skills like critical thinking, writing, and problem-solving because of AI. Teachers said some kids are relying on AI so much that they are not practicing spelling, having basic conversations, or working through hard problems on their own. Almost half of the teachers also said they do not support the government’s plan to give AI tutors to 450,000 lower-income students. They believe students who are already struggling need more support from real people, not just a chatbot. At the same time, many teachers are using AI themselves for work, and about half of schools still do not have clear rules for how staff or students should use it.

What parents and teachers need to know: Struggling through spelling, drafting a paragraph, or figuring out a tough question is often how learning sticks. Try this: after your child finishes an assignment with AI help, ask them to explain one part of it in their own words, no screen, no prompts. If they can, great. If they can't, that's your sign to continue to step in.

Inspired by this week’s section on AI literacy and the idea that kids should learn how AI works, not just how to use it, this game helps kids see that AI is not magic. It follows rules people make, and those rules can be fair, unfair, or missing important information.

What you need:
Paper, markers, and 10 to 15 small items like toys, blocks, crayons, snacks, or other safe household objects.

How to play

Why it works: This game helps kids understand that algorithms are not magic. They are just rules someone chose. It also gets kids thinking about fairness, missing information, and why humans still need to ask questions.

Even though there was less news this week, the AI train is not slowing down, and neither are we. Keep asking questions, keep talking to your kids and students, and keep reminding yourself that no algorithm replaces the messy, wonderful work you do every day. Catch you next Tuesday. Stay curious, and most importantly, stay human.

If this newsletter helped you feel even a little more grounded this week, then it did its job. If you know someone else who could use it, feel free to share it with them.

Thanks for reading, and I’ll see you next week.

Amber Ivey (AI)