Ofcom vows to take stronger action on platforms failing to take action to protect children online
Today Ofcom published the results of their call to action to tech platforms to update them on their progress on complying with the Online Safety Act’s children’s duties, identifying four themes: protection from grooming; safer content feeds; transparency on risk; minimum age enforcement.
We are pleased to see Ofcom taking a more robust approach by putting into the public domain the demands they have placed on platforms and the response they have received in return (including, notably, where platforms are falling short) as well as setting out the clear actions required from platforms in future and Ofcom’s own plans (see table on pp 6-7).
There is a welcome - and possibly overdue - threat that if services “are not complying with their duties to use appropriate systems to prevent children’s exposure to harmful content, we will proceed to enforcement action”. The regulator has to follow through on this, not least as their own research shows a shocking failure by sites to enforce their minimum age rules, with 84% of 8-12s reporting they had used one of the top five online services (YouTube, Facebook, TikTok, Instagram and Snapchat) in a month despite their minimum age requirement of 13+. This is not a marginal enforcement gap. It demonstrates that underage access is systemic across the platform ecosystem.
The shoddiness of platform response is underlined by the fact that, with the exception of TikTok, service providers did not respond to Ofcom’s invitation to explain to parents the steps they were taking to improve online safety. Ofcom suggests that this “represents a missed opportunity for companies to have demonstrated how seriously they take children’s safety” (p 7), or perhaps it is a recognition that too much transparency might not be good for business. In Ofcom’s words “when it comes to the largest sites – household names that children use the most – many have fallen short of putting children’s safety at the heart of their products.”
Equally concerning, given the OSA children’s duties have been in force for nearly a year, there has been little change in overall exposure to harmful content for 11-17s with around seven in ten (73%) of that age group saying they have seen or heard some form of harmful content online in a four-week period, comparable with Ofcom’s Wave 1 data. The data also highlights the central role played by recommender systems and engagement-driven design in exposing children to harm. Personalised feeds remained the single biggest route through which children encountered harmful content, with more than a third of 11 to 17-year-olds saying they saw harmful material while “scrolling on their feed”. This is a crucial finding as it shifts the focus away from isolated pieces of harmful content and toward the underlying design of platforms. The issue is therefore not simply whether harmful content exists online - it is that platform systems are actively surfacing, amplifying and distributing that content to children through algorithmic recommendation systems optimised for engagement.
Ofcom’s report notes that safer feeds for children were one of its key asks because of the likely impact on access to harmful content. Despite this, TikTok and YouTube did not commit to any changes on the basis that they think their feeds are already safe for children, and Meta committed to improving Facebook - to bring it in line with Instagram. Ofcom itself states that it is “deeply concerned by the responses to the safer feeds demands”, especially as regards TikTok and YouTube.
Ofcom’s findings on the time children have spent on platforms are also significant. Snapchat accounted for the largest share of children’s time online, with children in the highest-use quartile spending an average of 4 hours and 20 minutes per day on the platform. This level of engagement reflects the addictive and manipulative design features deployed on platforms such as Snapchat, including Snapstreaks, infinite scroll and disappearing messages, intended to maximise user attention and habitual use. This is a fundamental issue with platforms that are engineered around engagement metrics, ensuring that profits are always prioritised over user safety and wellbeing.
The failure to shift the dial in 11-17 year olds' exposure to harmful content starkly proves that despite immense public pressure, civil society criticism, platform assurances, and Ofcom’s implementation to date of the Online Safety Act itself, children’s lived experience online has not materially improved. The increase in children viewing bullying content, alongside their continued high exposure to hateful material, reinforces concerns that harm is structural. Ofcom’s latest research on children’s online experiences provides some of the clearest evidence yet that the Online Safety Act will only succeed if it is implemented through a genuine safety by design framework, not simply through ex-post content moderation or user empowerment tools, which are leaving children at risk.
As we have also called for, the OSA needs to be strengthened to ensure that Ofcom’s slow and cautious approach to implementation and enforcement - which itself is an issue here, as well as platforms’ willingness to comply with their rules - is improved. Yet, even if platforms had responded more positively to Ofcom’s demands for action, it is unclear whether these would have dealt with the problems in an unsafe environment - despite acknowledging the significance of the safer feeds demand.
Towards a stronger focus on safety by design
Yesterday, alongside 5Rights Foundation, Molly Rose Foundation, Internet Watch Foundation (IWF), End Violence Against Women Coalition (EVAW), Refuge and FlippGen, we launched a Safety by Design Code of Practice, which we are calling on the Government to mandate Ofcom to adopt. A safety by design approach would require tech platforms to conduct proper product testing to identify and mitigate foreseeable risks to children, and all users, at the point of development. This would mean that features and functionalities such as harmful recommender systems, dark patterns, autoplay, infinite scroll and viral amplification mechanisms are designed-out before they are rolled out onto the market.
Indeed, in a report by the BBC, Matt Navarra said the criticisms of platforms’ progress by Ofcom illustrated a shift to seeing online harms as “a product problem”, which is the point that we (and others) have been making for some considerable time. New polling published yesterday supports this assessment, with 84% of the UK public saying they are convinced that requiring companies to prove their products are designed to be safe before use would keep everyone safe on social media platforms.
Whilst harmful content remains widespread, underage access is prolific, and algorithm-driven feeds continue to facilitate exposure to harm, the central challenge for regulation is not simply platform compliance - it is platform design. We cannot expect blocking children’s access, or a ‘business as usual’ approach from Ofcom, to resolve these issues. Forty-four civil society organisations support amending the Online Safety Act to include a definition of safety by design, and an accompanying code of practice. The Government must use the opportunity it has through their consultation to take a more ambitious approach and ensure tech platforms are held to account for the continued harm they are causing by strengthening the existing regulatory regime.
We must demand a better and safer internet for all users.
