A new report from the Molly Rose Foundation (MRF) has revealed the extent to which ads are appearing alongside suicide, self-harm and depression content on TikTok and Instagram. The report comes as Ofcom begins enforcing the Online Safety Act in the UK, but uncovers “unacceptably pervasive and widespread” risk of exposure to harmful content on the two social media services.
The charity was founded by Ian Russell after his daughter Molly took her own life following exposure to harmful content on social media, and aims to tackle that content to protect young people online. Following comparable research in 2023, the MRF suggests that the risk profile on both platforms has not improved, and in TikTok’s case it has actively deteriorated. The findings raise “substantive questions about the likely efficacy of Ofcom’s current approach”, according to the report.
The advertiser’s role
The MRF analysed content across TikTok and Instagram between November 2024 and March 2025, including 644 short-form videos that were algorithmically recommended to accounts that were registered as a 15-year-old girl in the UK. It found the recommendation systems capable of “bombarding” young people with large amounts of harmful content. And in the analysis of algorithmically recommended videos featuring suicide, self-harm and depression content, 10 percent of the videos had ads before or after the harmful material.
While the report does not name specific brands whose ads appeared, it lists key groups and categories that showed up, including major fast-fashion retailers, major UK and international retailers, a large number of UK university and educational establishments, several major charities (including mental health charities), local youth-focused projects, and social media companies. The report urges companies to lean on social media services to ensure their ad spend is not inadvertently funding content that is harmful to children.
“Ensuring that children can no longer be algorithmically recommended harmful material is not only important for children and families, but it can also give advertisers and investors confidence that their adverts will no longer be shown next to high-risk material, nor that their ad-buying strategies contributing to a business model that exacerbates the risk of cumulative harm among young people,” says the report. “Advertisers each have a role to play to prevent social media continuing to cost young lives.”
The Instagram findings
While the study acknowledges that Instagram has taken measures to make it harder for young people to seek out harmful content, it found the Meta-owned service was still recommending “a near-constant supply of problematic suicide, self-harm and depression material.” On accounts that had previously been used to access this content, 97 percent of recommended videos on the Reels product were classified as harmful, thereby increasing the risk of “cumulative harm” to young users.
The report also suggests Instagram is “actively gaming” its requirements under Ofcom’s Children’s Code. The regulation requires the company to introduce a design feature enabling young users to provide negative feedback on content being recommended to them, but the feature also allows users to provide positive feedback and therefore be recommended similar types of content for the following 30 days, including suicide and depression-related material.
The MRF added that Meta’s investment in AI to drive more personalised recommendations could have a significant negative impact on children’s safety. “Unless strengthened and more effective regulatory safeguards can be quickly established, it seems difficult to conclude that regulatory incentives can reverse what many will conclude is yet another example of Meta pursuing a reckless rush for growth, with damaging implications for the under 25s,” says the report. “Safety, not solely profit, should be the guiding principle for the next generation of Meta’s content recommendation models.”
The TikTok findings
Meanwhile on TikTok, the social video platform continues to algorithmically recommend this kind of content in “appallingly large volumes”, according to the MRF, and has become more pervasive since the previous research in 2023. The charity found “no evidence at all” that TikTok has taken any measures to increase the platform’s safety-by-design, for example through introducing friction into the search and discoverability experience.
“TikTok’s For You page recommended a near continuous barrage of harmful material, with higher concentrations of the most severe types of material, including posts that promote and glorify suicide and that actively suggest suicide methods,” says the report. “We were more likely to be recommended more severe forms of content than on Instagram.”
And similarly to Instagram, the MRF found evidence that TikTok is seeking to game the Online Safety Act, by introducing features ostensibly designed to comply with the legislation, but again allowing users to provide positive feedback on suicide and self-harm related material.
The recommendations
The charity concludes that the companies “must be actively incentivised to shift away from an emphasis on user growth and commercial metrics towards a more balanced approach that emphasises both product safety and safety- and wellbeing-by-design.” They should also widen their definitions of harm towards a user-centered understanding of risk, where content is assessed by how, when and to whom it is being shown, rather than simply what it contains.
And Ofcom’s current framework is “clearly poorly placed to address the scale and magnitude of harm on major platforms”, according to the report, and the watchdog should revisit the scope and ambition of its regulatory approach. The report adds that the findings must also increase pressure on the UK Government to strengthen the Online Safety Act.
“Social media platforms must no longer be allowed to pay lip service to safety-by-design, while their designers and engineers remain actively incentivised to introduce increasingly insidious new engagement-based features that will inevitably deepen and extend the risk profile for young people,” says the report.
Follow VideoWeek on LinkedIn.
