Ofcom, the UK's communications regulator, opened a formal investigation into TikTok this week over whether its age-checking system is good enough to keep children away from harmful content. The investigation follows a broader review of how major platforms approach child safety, and the findings are uncomfortable for the whole industry, not just TikTok.
The specific concern is a technique called age inference. Instead of asking for a passport or running a face scan, platforms like TikTok watch how you behave: what you watch, who you follow, what words you use in your profile. They then guess whether you are a child. Ofcom says this guesswork may have failed to correctly identify a significant number of children, meaning those children ended up with access to content about self-harm, suicide, eating disorders, and pornography.
TikTok says it also asks users for a date of birth when they sign up and uses additional signals to catch underage users who lie about their age. It says it is confident it meets its obligations under the Online Safety Act and will cooperate with Ofcom's investigation. That confidence will now be tested.
The penalty structure is serious. Fines can reach £18 million or 10 percent of a company's total global revenue, whichever is higher. For a platform the size of TikTok, whose parent company ByteDance generates tens of billions in annual revenue, 10 percent would be an enormous sum. Ofcom can also apply to courts to have a platform blocked in the UK entirely in the most severe cases.
This matters well beyond TikTok. Ofcom made clear it has the same concerns about other platforms using age inference, and told them all to switch to stronger methods immediately. Instagram also uses this technique among others. The regulator's preferred alternatives are things like verified ID checks, facial age estimation from a photo or video, or confirmation via bank or mobile account data.
The broader picture shows the scale of the problem. Ofcom found that about one in ten teenagers aged 15 to 17 were still on the three most popular dating apps in December 2025, even after age checks were put in place on those apps. Meanwhile, more than half of results on Microsoft's Bing search engine led to adult content sites with no age checks at all, as did roughly one in three results on Google. Both search engines are now working with Ofcom to address this.
There is also a new front opening up around AI chatbots. Meta announced this week that it will now alert parents and, where appropriate, emergency services if a teenager discusses suicide or self-harm with the AI chatbot built into Instagram or Facebook. The announcement comes after legal claims from families of young people who died after conversations with AI chatbots, and after a Washington Post investigation found Meta's chatbot was coaching teen accounts on self-harm. Meta says parents who have signed up to supervision tools will receive these alerts in the UK, US, Canada, and Australia, with global rollout planned by year end.
All of this is happening against a fast-moving legal backdrop. The UK government announced on June 15 a full ban on social media for children under 16, covering TikTok, YouTube, Instagram, Snapchat, Facebook, and X. The ban is expected to come into force in Spring 2027. Australia ran a similar ban from December 2025, and early compliance data showed that while many underage accounts were removed, a substantial number of children still retained access.
The enforcement problem is the same everywhere: you can write the law, but the technical question of proving someone is under 16 without invading everyone's privacy remains genuinely hard. VPN usage in the UK surged dramatically after age checks came into force last July, including among children, which is precisely the group the rules were designed to protect.
For any business that runs an online platform, operates a community forum, sells age-restricted products digitally, or hosts user-generated content in the UK, the compliance bar is rising fast. Ofcom has already fined over 90 services and issued six formal fines since enforcement began. The TikTok investigation signals the regulator is now moving up the size ladder from smaller adult sites to the world's most used social platforms. The question of how you verify a user's age is no longer just a legal detail. It is a board-level risk.