TikTok and YouTube Fail Kids' Safety Test, Says

There's a particular kind of regulatory theatre that plays out whenever a government body confronts a tech giant: stern language, promised consequences, corporate PR in return. Ofcom's new report on children's online safety hits some of those familiar beats. But buried underneath the watchdog's rebuke of TikTok and YouTube is a more uncomfortable question—one that a group of schoolchildren from Greater Manchester articulated more clearly than most of the adults in the room.

The question isn't whether platforms are safe enough. They're not. The question is whether the tools governments are reaching for can actually do what they're supposed to do.

What Ofcom Actually Found

The headline finding is stark: nearly nine out of ten children aged 8 to 12 are still accessing platforms that are legally required to block them. That's not a rounding error. That's near-total failure of the current framework, almost a year after the UK's Online Safety Act came fully into force.

Ofcom's report doesn't distribute blame evenly. Roblox, Snapchat, and Meta received credit for introducing new anti-grooming protections. TikTok and YouTube, by contrast, were cited for failing to improve their content recommendation systems—which, if you've ever watched a child disappear into an algorithm for three hours, is where a lot of the real harm concentrates. YouTube responded that it provides "industry-leading, age-appropriate experiences" developed with child safety experts. TikTok called Ofcom's framing "very disappointing," saying the regulator failed to acknowledge its existing safety features.

Both companies' responses follow a template: acknowledge the concern, defend the product, imply the criticism is unfair. Whether that's spin or a legitimate grievance is worth holding open. Ofcom itself has faced criticism for being too tentative with tech firms, and the regulator now says it's prepared to use its full range of powers—including significant fines. That's a harder line than we've heard before. Whether it translates into enforcement is the test.

The Australia Problem

The most revealing thread in the BBC's reporting comes from an experiment Newsround ran: connecting Manchester schoolchildren with their Australian counterparts, six months after Australia implemented a ban on social media for under-16s. What the Australian kids described was instructive—and not in the way you'd hope.

"The ban seems a bit pointless," one young Australian said, "because if you're still going to be able to get round it, it's a bit like they just did all that work for nothing."

That's a fairly precise diagnosis of a policy gap. Australia's ban arrived without airtight age verification infrastructure—a structural problem the kids spotted immediately. One girl reportedly drew a fake moustache and scraped her hair back to fool face-scanning software. That's not a loophole; that's a barn door. And teenagers, as Newsround presenter Ricky Boleto noted with some sympathy, "always do find a way."

But Boleto also reported genuine upsides that don't get enough attention. Parks filling up. Kids talking to their parents more. Children who had previously been excluded—by parents who already restricted social media—suddenly finding themselves on the same footing as everyone else. "They could say, 'well, actually, none of us are on it,'" Boleto explained. The social levelling effect is real, even if the technical enforcement is porous.

This is the honest complexity of the Australian experiment: it hasn't worked cleanly, but it hasn't been meaningless either. Holding both of those things at once is harder than most policy coverage manages.

The Verification Gap Is the Whole Story

What ties Ofcom's findings to the Australian children's testimony is a single infrastructure problem: age verification doesn't work well enough to carry the weight being placed on it.

The UK law requires platforms to block under-13s using age verification tools. Nearly 90% of children in that age bracket are still getting through. Australia's ban extended that requirement upward to 16, without solving the underlying verification problem first. The education select committee in the UK is now calling for a statutory ban on social media for under-16s, plus restrictions on design features like infinite scrolling and disappearing messages for under-18s.

One Australian teenager framed the fix accurately: "If they put more thought into the verification things, that would have done them so much good, because that's how the majority of people have gotten around the ban."

She's not wrong. The policy architecture—ban first, verify adequately later—is backwards. You don't build a wall and then figure out how to make it solid.

Meta, for its part, has floated a different model entirely: shift responsibility to Apple and Google to lock down app stores according to device users' ages, creating a gatekeeping layer before a child even reaches a social media app. That's a genuinely interesting structural proposal, though it also conveniently relocates culpability away from Meta's own platforms. The idea that the operating system layer is where age enforcement should live has real logic to it—phones, after all, have more identity information than any individual app. But it raises its own questions about data, privacy, and who ultimately holds the verification key.

What the Consultation Might Actually Produce

Nearly 70,000 people participated in the UK government's public consultation on these potential measures, which is closing shortly. The government has pledged to act quickly on the results. What that action looks like will define whether the UK charts a more coherent path than Australia or repeats the same sequencing mistake.

The education committee's call for a statutory ban is the blunt instrument option. It generates political visibility and signals intent, but without better verification infrastructure it risks the same enforcement gap Australia is grappling with. The Ofcom approach—targeted pressure on platforms, specific requirements for recommendation algorithms, meaningful fines for non-compliance—is more surgical and potentially more durable, but it requires sustained regulatory will and the courts don't always cooperate on timelines.

There's a third dimension that gets underweighted in most of this coverage: the algorithm itself. A ban addresses access. Restrictions on infinite scrolling and disappearing messages address design. But neither gets directly at the content recommendation engine that, as Boleto put it, had him unable to stop watching Arsenal footage at midnight. If an adult journalist with full awareness of what's happening to him can't disengage, calibrating that same system for an eleven-year-old is not a small engineering challenge. It's the central one.

Let the Kids Testify

The most pointed observation from the whole BBC segment came from the anchor, almost as a throwaway: "always talk to the kids first. They're the ones with the ideas."

That's not just warmth—it's methodology. The Australian teenagers who described drawing moustaches to fool facial recognition, who complained that the ban lacked technical substance, who identified verification as the load-bearing flaw in the whole structure—they arrived at the sharpest analysis faster than most parliamentary committees. They're not neutral observers, obviously. They have interests. But they're also the most sophisticated users of these systems, and they understand the exploitation surface in ways that regulators often discover six months after the fact.

Ofcom, to its credit, is now pushing harder. The fines threat is real, and TikTok and YouTube face genuine exposure if they don't produce credible recommendation reforms. Whether that pressure produces meaningful change—or more technically compliant, effectively meaningless tweaks—is where attention should stay.

The children waiting to see what governments do next are watching with clearer eyes than most adults give them credit for. The parks being a bit fuller in Australia is a real thing. So is the moustache getting past the scanner. Both facts belong in the same sentence, because that's what the policy actually produced.

The question for the UK isn't whether to act. It's whether to act in a way that closes the gap rather than just drawing a new map of it.

Fatima Al-Hassan is Current Affairs Desk Editor at BuzzRAG.