In this latest episode of the Data Malarkey Podcast, Master Data Storyteller, Sam Knowles, spoke to Andy Burrows, Chief Executive of the Molly Rose Foundation. Their conversation covered online safety, platform accountability, and the continuing challenge of protecting children in the digital world.
The foundation was established following the death of 14-year-old Molly Russell in 2017. Since then, it has become one of the UK’s leading voices at the intersection of suicide prevention, online safety, regulation, and technology. Burrows joined the organisation in 2024 after more than a decade working on consumer protection and online safety, including with the NSPCC on the UK’s Online Safety Act.
The conversation centred on a difficult but important question: how much has changed since Molly’s death?
Plus ça change …?
Burrows described research conducted with the Conscious Advertising Network (CAN) that sought to answer precisely that. Keen-eared listeners will recall the first episode of Season Ten of Data Malarkey when this research was raised by our guest Harriet Kingaby, CAN’s Co-Chair. Her episode in all formats is available here.
Back to this critical study: researchers created social media accounts posing as 15-year-old girls and trained platform algorithms to recognise an interest in content relating to depression, self-harm, and eating disorders. The objective was not to recreate the experience of any specific teenager, but to test the safeguards supposedly designed to protect young users.
… plus ç’est la même chose
The results were deeply concerning. More than 95% of the algorithmically recommended content served to these accounts was classified as harmful, including material relating to suicide, self-harm, eating disorders, and severe depression. Burrows speaks powerfully about the cumulative effect of such content, describing an overwhelming environment of hopelessness and despair.
The discussion then moves beyond content itself to the economic incentives underpinning social media platforms. Burrows argues that these businesses are, first and foremost, advertising companies. Their commercial model depends on attention, engagement, and time spent on platform.
What are brands doing and what could they do differently?
A separate study by the foundation found adverts from household-name brands appearing alongside harmful content. This raises difficult questions not only about brand safety, but also about transparency. Burrows suggests that many advertisers may have little visibility into where their spending ultimately appears, or what content it may inadvertently help to monetise.
The conversation also explores the role of whistleblowers in exposing internal practices within technology companies. From Frances Haugen to Sarah Wynn-Williams, Burrows argues that such figures have helped illuminate decisions that would otherwise remain hidden from public view. However, he suggests that court disclosures may ultimately prove even more significant, providing direct evidence of how companies have responded to concerns about user safety.
Is social media having its “tobacco moment” in court?
One of the most striking sections of the discussion focuses on recent legal developments in the United States. Burrows describes court rulings against major technology companies as potentially significant because they shift attention away from individual pieces of content and towards product design itself – and in these most recent cases its addictive nature. The key question is increasingly not what users post, but whether platforms are designed in ways that knowingly create or amplify harm.
That distinction runs throughout the entire conversation. Burrows repeatedly returns to the idea that online harms are not inevitable consequences of technology. Rather, they emerge from business decisions, product decisions, and design decisions.
Should social media platforms be banned for tweens and teens?
The interview also touches on the UK’s Online Safety Act, proposals for social media bans for under-16s – accelerated in the embers of the Starmer administration since we recorded the podcast – and the challenges presented by generative AI. While Burrows acknowledges the complexity of each issue, his central argument remains resolutely consistent: meaningful change will not come from voluntary self-regulation alone. It requires stronger incentives, clearer accountability, and products designed with safety from the outset rather than added as an afterthought.
Perhaps the most memorable observation comes when Burrows recalled presenting evidence of increasing online grooming to a technology company, only to be told that such findings represented “today’s newspaper front pages and tomorrow’s chip-shop wrapper”. For him, that moment encapsulates the challenge.
The conversation ultimately leaves listeners with a sobering but important conclusion. Technology is not destiny. The systems that shape online experiences are built by people, governed by choices, and can therefore be changed. The question is whether society has the will to demand that change.
—
The first draft of this blog was written by ChatGPT, using a transcript of the episode and an ever-refined prompt. It was then edited by real humans.