The UK’s media regulator published a damning report today. YouTube’s response was to say everything is fine. One of them is wrong.
This morning, Ofcom published findings from its review of how the UK’s largest social media and video platforms are protecting children. The headline conclusion was stark: TikTok and YouTube have “failed to commit to any significant changes to reduce harmful content being served to children, maintaining their feeds are already safe for children.”
Ofcom’s position, backed by its own research, is that they are not. The regulator found that 84% of children aged 8-12 are still using at least one major platform with a minimum age of 13 – and that the feeds those children are being served remain the primary route through which they encounter harmful content online.
YouTube’s response was to point to existing features – timers, age-appropriate experiences, work with child safety experts. TikTok called the criticism “very disappointing.”
Neither response engaged with the question that matters most.
The debate has shifted – and the platforms haven’t kept up
Social media consultant Matt Navarra, speaking in response to today’s report, put it precisely: “The old debate was, ‘did the platform remove harmful content quickly enough?’ – the new one has shifted towards, ‘why did the platform show it to a child in the first place?’”
That shift is significant. For years, the conversation about children’s online safety has been framed around content moderation – whether platforms were fast enough to remove videos that violated their policies. YouTube’s existing defences make sense within that old frame. It has policies. It has human reviewers. It removes content that breaks the rules.
But the Ofcom report is pointing at something different. The problem isn’t only that harmful content exists on YouTube. It’s that YouTube’s recommendation algorithm actively decided to show it to a child. The feed isn’t a passive library that children browse. It’s a personalised, constantly-updated selection of what the platform calculates will keep that specific child watching for as long as possible.
Content moderation can flag what’s already known to be harmful. It cannot anticipate what the algorithm will serve next – and it cannot change the fact that the algorithm’s objective is engagement, not wellbeing.
What YouTube’s safety features actually address
YouTube’s response to Ofcom cited its “industry-leading, age-appropriate” experiences – including tools that let parents set time limits on the Shorts feed. These are real features, and they’re not nothing.
But they address the wrong question. A timer tells a child when to stop watching – but not why stopping is so hard in the first place. It says nothing about what they watched, or why YouTube chose to show it to them. Age-appropriate content levels reduce the pool of material the algorithm draws from. They don’t change the fact that the algorithm is still running, still optimising for engagement, and still deciding – without any parental input – what plays next. We’ve examined what those age-appropriate content levels actually do – and don’t do – in detail here.
Ofcom chief executive Dame Melanie Dawes, speaking on Radio 4’s Today programme this morning, acknowledged the scale of what’s being asked for. “We’re talking about a twenty-year culture at Silicon Valley of not taking safety seriously,” she said. “You can’t change that overnight.”
She also confirmed that Ofcom is “ready to take the toughest enforcement action” if platforms fail to comply – including formal investigation. The Education Committee, separately publishing its response to the government’s social media consultation today, called the situation plainly: “Social media firms cannot be relied upon to self-regulate.”
The answer the report points towards – but can’t mandate
Andy Burrows, chief executive of the Molly Rose Foundation, welcomed the report but went further, calling for “a conditional ban on personalised algorithms that continue to push out a tsunami of harmful content to teens.”
That’s the logical endpoint of the Navarra framing. If the algorithm is the mechanism through which children encounter harm, and if content moderation cannot adequately police what an algorithm will recommend next, then the algorithm itself is the thing that needs to be addressed. This is the process researchers call algorithmic creep – and it operates regardless of whether any individual piece of content violates a platform’s policies.
Regulation may eventually get there. Legislation takes time. In the meantime, parents cannot wait for Silicon Valley to change its twenty-year culture, or for Ofcom’s enforcement timeline to run its course.
What parents can do right now
The question Ofcom is now asking – why did the platform show it to a child in the first place? – has one answer that doesn’t depend on YouTube’s compliance or regulatory timelines: remove the algorithm from the equation entirely.
Whitelisting means choosing, in advance, exactly which channels a child can access – and making everything else inaccessible. The algorithm never runs because there’s no opportunity for it to run. There is no feed to moderate, no recommendation to second-guess, no next video the platform has chosen. There are only the channels a parent has reviewed and approved. It also means AI-generated content farms – one of the most prevalent forms of low-quality content the algorithm serves to children – can never reach them.
It doesn’t solve every problem the Ofcom report identifies. It doesn’t address grooming risks on messaging platforms, or age verification failures across social media. But for YouTube – the platform that came second only to TikTok in Ofcom’s harmful content findings, and the one most young children use daily – it is the only approach that addresses the “why did it show it?” question at its root. See our guide to the full range of YouTube safety options available to parents right now.
The feed is the problem. Streamu removes the feed.
Streamu lets parents build their child’s YouTube experience from a curated list of approved channels – with no algorithm, no autoplay, and no route into the wider YouTube ecosystem. What your child watches is what you’ve chosen. Nothing else.
Join a growing community of conscious parents taking back control.
Join the waitlist at streamu.app
Sources: Ofcom children’s online safety report, 21 May 2026 (BBC News); TikTok and YouTube lag on UK child safety as rivals act (Reuters via Rappler)
