When Cybersmile launched Stop the Forced Feed, we challenged one of the internet's most powerful assumptions: that online safety could be achieved simply by removing harmful content after it appeared. We have always believed that the conversation had to go further.
The recommendation systems responsible for amplifying, prioritizing, and repeatedly delivering harmful content into people's feeds also had to become part of the solution. That principle became the foundation of Cybersmile's Algorithmic Harm Initiative—combining research, campaigning, and policy engagement to tackle one of the internet's fastest-growing online safety challenges.
From campaign to regulatory milestone
In 2025, Cybersmile conducted one of its most significant pieces of research to date into algorithmic harm, published its findings, and submitted that evidence directly to Ofcom.
Ofcom has since announced sweeping new measures requiring technology companies to take greater responsibility for the role recommendation systems play in exposing children to harmful content online. Under the new rules, platforms must assess how their algorithms create risk, prevent harmful content from being recommended to children, and redesign systems that place young users in harm's way.
For Cybersmile, Ofcom's announcement represents another landmark victory in our fight against algorithmic harm—and another significant milestone in the journey that began with Stop the Forced Feed. It also marks the second major external milestone for the campaign in recent months.
Earlier this month, Meta announced significant changes to Instagram's recommendation systems. CEO Adam Mosseri introduced expanded topic controls, greater transparency around recommendations, and new ways for users to influence the content they see—developments that closely reflected many of the same principles that Cybersmile has consistently championed through Stop the Forced Feed. Taken together, these developments tell a powerful story.
Across both industry and regulation, the conversation is changing. Recommendation systems are no longer viewed simply as tools that maximize engagement. They are increasingly recognized as powerful digital environments that shape people's experiences and therefore carry significant responsibilities for user wellbeing and online safety. This is precisely the conversation Stop the Forced Feed set out to lead.
Stop the Forced Feed: The journey so far
Campaign launch
Cybersmile launched Stop the Forced Feed, challenging the role of recommendation systems in online safety and introducing algorithmic harm as a defining online safety issue.
Research & evidence
Cybersmile conducted one of its most significant pieces of research into algorithmic harm and submitted its findings directly to Ofcom, the Commissioner for Human Rights, the European Commission, and the British Government.
Regulatory milestone
Ofcom introduced new measures placing legal responsibilities on platforms to assess and reduce the risks created by recommendation systems.
Industry milestone
Meta introduced expanded topic controls, greater transparency, and increased user agency across Instagram's recommendation systems, closely reflecting many of the principles Cybersmile has consistently championed through Stop the Forced Feed.
Building momentum
Together, these milestones demonstrate growing recognition across industry and regulation that recommendation systems themselves have become central to creating safer online experiences.
Why algorithmic harm matters
From the outset, the campaign argued that harmful online experiences are influenced not only by the content people encounter but also by the algorithms deciding what is promoted, amplified, and repeatedly delivered into their feeds.
Our research into algorithmic harm demonstrated how engagement-driven recommendation systems can unintentionally increase exposure to harmful material within seconds, often without users understanding why content is appearing or having meaningful control over it. Those findings formed the basis of the evidence Cybersmile submitted directly to Ofcom, helping place algorithmic harm firmly within the regulator's evidence-gathering process.
Today's announcement demonstrates just how far that conversation has progressed. For years, online safety discussions focused primarily on moderation—how quickly harmful content could be identified and removed. While moderation remains essential, it addresses only part of the challenge.
If harmful material continues to be recommended, amplified, and repeatedly surfaced by engagement-driven systems, removing individual posts alone will never fully address the underlying risk. That is why algorithmic harm and user agency matter.
When we launched Stop the Forced Feed, we believed online safety had reached an important turning point. Harmful content doesn't simply appear in people's feeds by chance; it is often delivered there by recommendation systems designed to maximize engagement. We argued that if we wanted to build a safer internet, we had to address algorithmic harm itself. Today's announcement is another landmark milestone in that journey. It demonstrates how far this conversation has come and reinforces the direction Cybersmile has consistently advocated from the very beginning.
Scott Freeman, CEO, The Cybersmile Foundation
A defining moment for online safety
The significance of today's announcement extends well beyond social media. Recommendation systems increasingly influence the information we consume, the communities we discover, the opportunities we encounter, and the decisions we make. As artificial intelligence becomes embedded across more aspects of everyday life, ensuring these systems operate responsibly will become one of the defining challenges of the digital age.
The research and evidence Cybersmile submitted directly to Ofcom on algorithmic harm now sits alongside a regulatory framework that places recommendation systems at the heart of protecting children online. That is a significant moment—not only for Stop the Forced Feed but also for the wider evolution of online safety.
In the space of just a few months, Cybersmile has seen two major external milestones reflect the direction the campaign has consistently championed. Firstly, one of the world's largest technology companies introduced significantly greater user control over recommendation systems, and secondly, the UK's online safety regulator has placed new legal expectations on the algorithms responsible for recommending harmful content to children.
Together, these developments represent something much more significant than two isolated announcements. They demonstrate growing recognition across both industry and regulation that recommendation systems themselves have become a central online safety issue. Algorithmic harm is no longer a niche concern discussed only by researchers and campaigners. It is rapidly becoming a defining issue for governments, regulators, and technology companies alike.
For Cybersmile, that is both encouraging and deeply significant. Stop the Forced Feed was never simply about changing one platform or influencing one policy. It was about reshaping how society thinks about online safety.
Stop the Forced Feed was founded on a simple belief: that if we wanted a safer internet, we had to address not only harmful content but also the systems responsible for delivering it. Recent developments across both industry and regulation suggest that conversation is no longer emerging—it is becoming established.
The journey continues
Our work around algorithmic harm and user agency is not complete, but the direction of travel has never been clearer. The future of online safety will not be defined solely by the content platforms remove. It will also be defined by the algorithms they choose to build.
What began with Stop the Forced Feed has grown into a much broader movement towards greater accountability for the systems shaping our lives. Cybersmile will continue working with industry, regulators, and policymakers to ensure that momentum becomes lasting change.