Happy Thursday everyone - but maybe not such a happy one for Meta.

A judge in Santa Fe, New Mexico has started hearing arguments about whether Instagram and Facebook should be ordered to remove infinite scroll, push notifications and the algorithm itself for children. Which would be enormous. Meta's response: if the judge orders all of it, it may not be able to comply and the company might just pull its services from New Mexico entirely. So that's where we are. Either redesign the products, or take them away.

And then on Tuesday, Meta announced it is rolling out AI to hunt down the under-16s who got onto Facebook and Instagram by lying about their age. Which is, to be fair, a bit of an admission. Millions of children have been on these platforms with adult birthdays attached to their accounts, slipping past the protections meant for teens, and now the company has built a tool to find them. The phrase "closing the stable door after the horse has bolted" comes to mind, except in this case the company built the door, opened the door, and is now trying to find the horse.

The UK has chimed in too. The government has quietly taken on the legal power to restrict the same product features the New Mexico court is being asked to ban — infinite scroll, push notifications, the algorithm — without needing a new Act of Parliament. Different route. Same target.

Plenty to dig into, so grab that coffee.

— Heidi

On 5 May, Meta announced it is expanding an AI tool that detects suspected under-16s on its platforms — even on accounts that registered with an adult birthday. It is now live on Facebook in the United States and on Instagram across 27 EU countries plus Brazil. UK and EU expansion to Facebook is scheduled for June.

The AI scans entire profiles for the things teenagers do without thinking: birthday posts, comments wishing someone happy birthday, photos with cake, mentions of school. When it flags an account as likely a teen, the account is automatically moved into Teen Account protections. Private by default. Tighter content filters. Messaging from strangers blocked. Meta says millions of accounts have already been moved through this method.

If they want to override the restriction, they will need a parent's involvement. The "just enter a different birthday" option is gone for accounts the AI has flagged.

What to do this week:

Check what age the account is registered as. Open Instagram or Facebook on your child's phone. Settings → Account → Date of birth. If it shows an adult age and your child is under 18, the safety defaults that should have been active have not been.

Don't wait for Meta to catch it. Quiet teen accounts can slip through the AI. And if the AI does flag the account, your child finds out before you do. The conversation about why their account has just been restricted is harder than the conversation about fixing the date of birth together this week.

Decide whether you want the account in place at all. The platform has just admitted, through its actions, that millions of teen accounts have been operating without the safety protections that were supposed to apply to them. If you have been on the fence, this is your prompt.

Read the full piece: Meta is now using AI to catch kids who lied about their age →

Two pieces of legislation got Royal Assent on 29 April. They didn't make a lot of noise. They matter.

The Children's Wellbeing and Schools Act gives the Secretary of State the power to introduce age-based or feature-based restrictions on children's social media — through secondary legislation, meaning regulations can be made quickly without a new Act of Parliament. Ministers have signalled they want the first restrictions in place by the end of 2026.

The same Act made the government's mobile phone ban guidance for schools statutory. It was advisory. Now it is law. Every school in England is now legally required to follow it. Teachers and headteachers have been asking for this clarity for two years, and now they have it.

The Crime and Policing Act, which also got Royal Assent on 29 April, separately criminalises the creation and supply of so-called "nudification tools" — software designed to generate non-consensual intimate images of real people. It also brings more AI chatbots within the scope of the Online Safety Act. Both have been used against children, and the law is now catching up.

What to do this week:

If your child is at school in England, the phone rule is now legal, not advisory. That changes the conversation if your child has been pushing back. The school's hands are no longer tied.

If you want a say in what the social media restrictions look like, the consultation closes 26 May. Ten minutes. Search "Growing Up in the Online World consultation" on gov.uk.

Read the full piece: The UK now has the power to restrict children's social media →

The trial began on 4 May 2026 in Santa Fe. Judge Bryan Biedscheid is hearing arguments about whether Meta should be ordered to fundamentally redesign Instagram and Facebook for children — not fine the company, not issue a warning, but redesign the products themselves.

This is phase two of a case that started in March, when a New Mexico jury found that Meta knowingly harmed children's mental health and concealed what it knew about child sexual exploitation on its platforms. The jury ordered Meta to pay $375 million in civil penalties. Now prosecutors are asking the judge to order what Meta must actually change.

What the state is asking the judge to order:

Meta's response in court has been blunt. The demands are "overbroad, vague, unworkable." If the judge grants them in full, the company has said it may cease offering its platforms in New Mexico altogether.

What happens in Santa Fe will not stay there. A federal trial involving over a thousand US school districts begins on 15 June and rests on the same legal theory. The features being argued about (infinite scroll, push notifications, the algorithm itself) are the same ones regulators in the UK, Australia, and across Europe are trying to restrict.

What this means for you right now:

The trial will run for about three weeks. Nothing changes for your child's account today. But the features being argued about in court are ones you can address yourself, in settings, today.

The court will decide what Meta must do. You don't have to wait for the verdict to do some of it yourself.

Read the full piece: A judge is deciding whether to force Meta to redesign Instagram for children →

If you work in fintech or finance, you already have too many tabs open and not enough time.

Fintech Takes is the free newsletter senior leaders actually read. Each week, I break down the trends, deals, and regulatory moves shaping the industry — and explain why they matter — in plain English.

No filler, no PR spin, and no “insights” you already saw on LinkedIn eight times this week. Just clear analysis and the occasional bad joke to make it go down easier.

Get context you can actually use. Subscribe free and see what’s coming before everyone else.

Sign Up Free

Apple has quietly added a setting that does what fake birthdays defeat. The Declared Age Range API went live worldwide on iOS 26, iPadOS 26 and macOS 26. If your child uses a Child Account through Family Sharing, you can set their age range in Screen Time, and supporting apps receive that signal automatically when your child opens them. Set it once, every supporting app reads it. The list of supporting apps is short for now and growing. Worth five minutes of setup this weekend.

Roblox is splitting under-16 accounts into two age groups in early June. Children aged 5–8 will be moved into "Roblox Kids" accounts, where chat is off by default and game access is narrowed to titles that have passed an additional review. Children aged 9–15 land in "Roblox Select" with filtered games up to a Moderate rating. Both are stricter than the current defaults, and the chat-off default for the youngest age group is the bigger of the two changes. Check the date of birth on your child's Roblox account before June. If it's wrong, the safety defaults won't match their actual age, and they may lose access to games they have been playing for years.

Meta is now using AI to find under-16s who lied about their age when they signed up. The simplest thing you can do this week is check the date of birth on your child's account before the AI does it for you.

Ofcom is expected to publish its child safety compliance report on the six major platforms — Facebook, Instagram, TikTok, Snapchat, YouTube and Roblox — sometime in May. The platforms had until 30 April to respond to four formal demands. If Ofcom is not satisfied, fines of up to £18 million or 10% of global turnover are on the table.

The UK consultation on children's social media access closes on 26 May.

The New Mexico trial continues into its second week.

Until next Thursday!

— Heidi