Happy Thursday — and if you're already in the school holidays, I hope it's going well, screen-free, screen-full, or somewhere in the middle!

This week, Apple did a thing. For years, the honest answer to almost every new app, device or feature was: good luck, you're on your own. Something would launch, land in your child's hands, and it was left to parents to work out whether it was safe, usually after the fact. That's been the default for so long it started to feel permanent. But this week Apple finally cracked, and what they've done is the closest thing to "we'll help" parents have had in a long time. Nothing to do just yet. Find out more below.

Canada became the first country to put real legal duties on AI chatbots, not just social media. And a neuroscientist argued, loudly, that schools handed too much over to screens too fast.

Are governments, tech companies and researchers finally treating it as their problem too?

So, grab your coffee — and I'll fill you in on what's been happening this week.

— Heidi

Apple has previewed its biggest set of parental-control changes in years, and several are things parents and safety experts have wanted for a long time. The two that matter most: children will need a parent's approval before opening a new website in Safari, and you'll be able to set screen-time limits by category — games, social media and entertainment each getting their own allowance — instead of one blanket limit.

There's also a redesigned Screen Time dashboard that shows where a child's time is actually going and lets you pause access in a tap, and a simpler setup that starts a child's device with a handful of essential apps rather than a fully loaded phone. Communication Safety, which already blurs nude images, will start catching violence and gore too, and stays on by default for under-18s.

One catch worth knowing: this is a preview. The features arrive with Apple's software updates in the autumn, so there's nothing new to switch on today. The most useful thing you can do now is make sure your child's device is actually on a child account — everything Apple announced builds on top of that, and the protections that already exist are working today.

Read the full rundown: Apple just gave parents the controls they've been asking for →

For two years, the policy fight over children and tech has been about social media: age limits, bans, verification. This week it widened. When Canada introduced its online safety bill on Wednesday, it didn't only propose barring under-16s from social media. It placed legal duties on AI chatbots too, requiring them to handle crisis situations, avoid steering children towards harm, and reduce the risk of harmful behaviour. It appears to be the first time a national government has written AI chatbots into a children's online safety law this directly.

Canada isn't moving alone. Over the past year a string of US states has begun regulating "companion" chatbots, the apps built to act like a friend or a confidant. California's law, in force since January, requires these bots to tell users they aren't human, remind under-18s every few hours that they're talking to a machine, block sexual content for minors, and step in when a conversation suggests self-harm. New York and Washington have passed their own versions, with Washington's letting families sue directly. Idaho and Oregon now bar chatbots from claiming to be sentient or starting sexual conversations with children. By early this year, more than thirty states had bills in motion.

The reason is sobering, and worth being plain about: these laws exist because children have been harmed. The case that drove California's law was a fourteen-year-old who died after forming an intense attachment to a companion bot. Surveys suggest most teenagers have now used one at least once.

None of this is something to change in a settings menu today. The rules are still arriving, and most apply to the companies rather than to you. But it's a useful prompt. If your child uses an AI companion app, it's worth knowing which one, what it's built to do, and how it responds — because the guardrails these laws now demand are precisely the ones many of these apps have been missing.

We've just updated our full guide with where the rules now stand: AI Chatbots and Children →

The professionals pulling ahead aren't working more. They're using Claude.

Our free guide will show you how to:

Transform your workflow with AI and stay ahead of the curve with this comprehensive guide to using Claude at work.

Get The Free Guide

Where the world is moving on children's online safety law. Full picture on our live tracker →

🇨🇦 Canada introduced the Safe Social Media Act, its long-awaited national online safety bill. It would bar under-16s from social media — with an exemption for platforms that can prove they keep children safe — and, unusually, place legal safety duties on AI chatbots as well. It's only at first reading, so there's a long road ahead, but a G7 country moving is a significant marker.

🇬🇧 United Kingdom confirmed its school-phone guidance becomes legally binding from 29 June, with Ofsted now assessing every school's phone policy at inspection. Guidance schools could take or leave becomes something they have to act on.

🧸 The AI "companion" toys are back in the gift guides. Researchers who tested six chatbot-powered soft toys this month found the same problems as before: human-sounding flattery that builds attachment in very young children, "infinite chat" features funnelling what kids say into data pipelines, and almost no age checks. If you're weighing one up, our guide walks through what to look for: Are AI Toys Safe for Children? →

🏫 A neuroscientist's case against classroom tech. In The Digital Delusion →, Dr Jared Cooney Horvath argues that EdTech has quietly made children's learning worse, not better. We wouldn't sign up to every claim in it — a couple of the headline stats don't survive a close look — but the questions it hands parents to put to a school are genuinely useful: Is this deliberate? Is it time-limited? Is it justified? And can you show me the evidence it helps?

🍎 Apple's new controls arrive with its software updates in the autumn. We'll walk through setup when they land.

🌍 More bans incoming. Denmark and Austria are both expected to publish draft legislation this month — under-15 and under-14 respectively.

🎮 Roblox's legal pile keeps growing, and a US trial representing school districts is set to begin. Worth watching for the first real verdict.

Till next Thursday....

— Heidi