Source: Ofcom published on this website Thursday 21 May 2026 by Jill Powell
Children in the UK will be better protected from harm as some of the largest and most popular sites and apps agree to introduce significant new anti-grooming measures in response to Ofcom’s public call for action.
Snap, Meta and Roblox have confirmed to the regulator that they will bring in new safety measures designed to better protect children from online strangers who wish them harm - ranging from tighter default settings for children’s contacts and friendship groups, to AI detection tools and direct chat controls.
These new commitments are included in a report, published today, which confirms how major platforms – Facebook, Instagram, Roblox, Snap, TikTok and YouTube – have responded to Ofcom’s demands in March to urgently strengthen protections for children online.
Under the new commitments secured for UK users, children’s safety will be considered upfront by platforms before new features are rolled out to them. In response to our call to end product-testing on children, five of the services have committed to notify the regulator whenever they update their risk assessments before making significant changes to their services – going above and beyond their duties under the Online Safety Act.[1] This will ensure that Ofcom is aware of new products and features before they launch, allowing us an opportunity to scrutinise any adverse impact on children.
Despite these important improvements, there are areas where the response from tech companies raises serious questions. We are particularly concerned about the lack of action from some platforms to make their feeds safer. Similarly, we are not convinced that the commitments from platforms with a strict minimum age of 13 will effectively prevent underage children from accessing their services.
Given these outstanding concerns, we have today announced a five-point action plan to drive further change and hold platforms to account for the safety of children on their platforms.
Online grooming can have devastating, lifelong consequences for victims and we demanded action from tech firms to introduce failsafe online protections for children against adult predators.
In response, Snap has agreed to make significant and long-overdue changes by adopting all the recommended grooming prevention measures under our Illegal Harms Codes. As a result, adult strangers will be prevented from contacting children on Snap by default, and children will no longer be encouraged to expand their friendship groups to people they don’t know. These protections apply only to children using Snap in the UK, and mean they will have stronger safeguards against grooming than users in other countries. The platform will also roll out highly effective age-checks to all users over the summer to ensure that all under-18s in the UK benefit from these new safety measures.
Roblox has committed to build on its existing anti-grooming protections, announced earlier this year.[2] The company will now go even further, including by giving parents the ability to switch off direct chat services entirely for under-16s.
Meta has committed to developing a new setting that will hide teens’ connection lists on Instagram by default. The company also plans to roll out AI tools to detect likely sexualised conversations between adults and teens in Instagram direct messages, to report offending accounts to the NCMEC, and to take necessary enforcement action.
We are clear that these commitments must now translate into action. We’ve set clear deadlines by when we expect to see these changes in place and will be scrutinising how effectively they are implemented. We also want to see Meta go even further to prevent grooming and are pushing the company to adopt the network expansion prompts safety measure set out in our Illegal Harms Codes.[3]
If these promised improvements happen too slowly, or are not properly implemented, we will not hesitate to act.
Content feeds are children’s primary pathway to harm and we are deeply concerned by the overall response from industry to our challenge to fix the problem.
Meta pledged to extend its 13+ “movie‑style” content settings currently available on Instagram to Facebook, which it says will make its feeds safer by limiting the content teens see to a more age-appropriate level. While this is encouraging, it is too early to tell whether this will meaningfully reduce harm in practice.
Notably, TikTok and YouTube failed to commit to any significant changes to reduce harmful content being served to children, maintaining their feeds are already safe for children. Our wealth of evidence, published today, suggests they are still not safe enough.[4]
Since the children’s online safety duties came into force in July 2025, there has been little change in children’s overall exposure to harmful content, with nearly three-quarters of 11- to 17-year-olds (73%) encountering it in a four-week period. Just over a third (35%) of these children recalled exposure to harmful content when they were ‘scrolling on their feed’. Half of secondary school aged children who see harmful content recall coming across it on TikTok (53%), followed by YouTube (excluding YouTube kids) (36%), Instagram (34%), and Facebook (31%).
We are sharing these concerns with government as it concludes its consultation: Growing up in the online world.[5]
We have already issued in-depth, legally-binding requests for information to Meta, TikTok and YouTube on how they detect and prevent children being exposed to harmful content and are currently interrogating their responses. In order to independently verify what services are telling us, we are exploring using new inspection powers under the Online Safety Act. This would require firms to undergo an independent audit - carried out by a skilled expert in this field - and allow us to issue remote inspection notices to observe how services’ content detection, moderation systems, algorithms and age-checks are working in real time.
We are determined to drive change by all means necessary, unless services can demonstrate urgent and credible action plans to make feeds safer for children. Any service that ultimately fails to comply with their duties to protect children, can expect enforcement action.[6]
Tougher legislation to underpin minimum age enforcement
Despite acknowledging the importance of minimum age policies, none of the companies with a minimum age of 13 on their services convinced us that they are currently enforcing them effectively and the impact is clear. Our latest research shows that 84% of children aged 8–12 are still using one of the top five reaching online services (YouTube, Facebook, TikTok, Instagram and Snapchat) despite a minimum age of 13.[7]
Current online safety laws do not explicitly require services to keep underage children off their platforms by using robust age checks, although the Information Commissioner’s Office can take action in this area under data protection law. [8] We have today written to the Secretary of State to advise that, should Government and Parliament wish Ofcom to be able to force firms to enforce minimum age policies effectively, this would need a clearer basis in online safety legislation.
Ofcom’s public scrutiny of these six online services will continue under our five-point action plan. We will continue to drive change by:
- Monitoring the implementation of services’ new commitments to ensure they work in practice to better protect children.
- Continuing to review evidence on recommender feeds and explore whether to use new inspection powers, with a further update in July.
- Launching enforcement action where we have grounds to suspect non-compliance with the Online Safety Act.
- Continuing to monitor the real-world experiences of children online through our world-class research.
- Engaging with Government on the outcome of its consultation so that enforcement of minimum age policies under online safety laws can be strengthened where needed.
