Keeping children safe online: what we learned at Labour Party Conference 2026

2026 has been a defining year for online safety in the UK. It has been a year marked by a wave of government announcements and tech controversies, including the emergence of AI-generated sexually explicit imagery and proposals for restricting social media access for under-16s. What is clear, is that the risks facing children online have never been more urgent and complex.

Against this backdrop of heightened activity, alongside renewed cross-party calls for stronger AI safety regulation, the Labour Party Conference was a timely opportunity to assemble a panel of experts to assess the state of play and consider what action the UK Government must take to make a meaningful difference to children’s online safety.

Our panel featured the Rt Hon Ian Murray MP, Minister of State for Creative Industries, Media and Online Safety; former safeguarding minister Jess Phillips MP; Catherine Reader of FlippGen; and Kerry Smith, Chief Executive of the Internet Watch Foundation (IWF).

The state of play on online harms facing children

Kerry Smith opened by setting out what the IWF is seeing on the frontline. Last year, the IWF identified record levels of child sexual abuse material online, taking action on more than 311,000 reports containing confirmed illegal abuse material.

AI is accelerating the problem, making increasingly realistic abusive imagery possible at unprecedented speed and scale. The IWF’s warning was clear: this is not a future threat; it is happening now, and at pace.

Catherine Reader brought young people’s experiences into the discussion. Young people rely on platforms such as TikTok for education, entertainment and connection, but can also encounter harmful and distressing content. The same online spaces that help them connect can quickly become toxic.

Jess Phillips MP reflected on how dramatically the landscape has changed. Her “greatest regret”, she said, was that she and fellow safeguarding campaigners did not fully understand a decade ago the scale of what was coming.

Children are now exposed to harmful sexual behaviours and attitudes in ways previous generations were not, alongside growing concerns about harmful behaviour among young people themselves.

Will the UK social media restrictions be enough to protect children?

The Government’s proposed restrictions on social media use for under-16s were a central part of the discussion.

The Minister described a blanket age ban as a “blunt tool”, arguing instead for a broader approach focused on the features and functions that create risk. Under the proposed model, services could face restrictions unless they demonstrate that unsafe features – including infinite scrolling, recommender algorithms, addictive design and communications with strangers – have been removed or adequately mitigated.

A policy designed to protect children on one part of the internet cannot succeed if it simply pushes harmful activity somewhere that is harder to detect and disrupt.

The Minster asserted that this safety-led model would preserve an incentive for companies to improve their products. For example, platforms could earn an “off-ramp” from restrictions by proving that their services were safe. It would also place the burden where it belongs, requiring companies to demonstrate safety to the regulator rather than expecting children and families to absorb the risks of harmful design.

The Minister was equally direct about resistance from industry. Acknowledging that the Government was likely to face legal challenges, he said: “We’re going to end up in court … but bring it on.” Pointing to litigation against Meta in the United States and settlements relating to harm to children, he argued that technology companies have had too long to address foreseeable risks. The prospect of litigation, he made clear, must not dilute effective protections.

This is why the IWF has consistently argued for a coordinated, multi-layered approach rather than relying on any single intervention. Our recent recommendations to Government cover end-to-end encrypted environments, AI chatbots, device-level protections, and strengthening the Online Safety Act.

Preventing displacement into private and encrypted spaces

Encryption provides important privacy and security benefits. But where safeguards are absent, end-to-end encrypted environments can also become spaces where child sexual abuse material is shared and where children can be groomed.

Kerry highlighted the vulnerability of children whose intimate images are being shared within private groups. Removing an image from the open web is one thing. Removing it from the private messaging groups where a child's peers are seeing and sharing it is another.

The challenge, therefore, is not to choose between privacy and safety. It is to develop solutions that protect both.

Jess Phillips MP also challenged the technology industry’s use of privacy and freedom of speech as arguments against measures such as encryption regulation, age verification and device-level safeguards.

She argued that the largest technology companies already collect vast amounts of personal information about users, from location and purchasing habits to highly sensitive data. In her view, this makes claims that stronger child-safety measures would fundamentally undermine users’ privacy difficult to reconcile with the existing data practices of major platforms.

Her broader point was that privacy and free speech should not be used as blanket arguments against measures designed to improve children’s online safety, particularly when the companies making those arguments already have extensive access to users’ data and significant influence over online expression.

This is an area where technology already has a role to play. The IWF has highlighted privacy-preserving approaches such as upload prevention, which can detect and block known child sexual abuse material before it is encrypted.

Regulation only works when it is enforced

The discussion also turned to the Online Safety Act and Ofcom. There was recognition that the Act provides an important regulatory framework, but that implementation and enforcement need to keep pace with the technology and the harms being addressed.

The Minister acknowledged concerns about Ofcom's pace and effectiveness, arguing that the regulator may sometimes need to act more quickly and decisively.

The scale and pace of the legislative task underline the importance of getting enforcement right. The minister said that two substantial pieces of primary legislation, alongside secondary regulations, would need to be passed by December ahead of the restrictions coming into force early next year. The regime would then be refined through rules addressing particular features and functions, with regulations for AI chatbots to follow. Enforcement, he acknowledged, will need to be strict, adaptable and prepared for a continuing “whack-a-mole” challenge as services and harms evolve.

AI is both a challenge and an opportunity

AI ran through the discussion as both a growing threat and a potential part of the solution.

The technology is already being misused to generate increasingly realistic child sexual abuse material. But AI can also help identify harmful content and detect patterns of abuse. The challenge is ensuring that safety is built into AI systems from the start, rather than added after harm has occurred.

Our recent work has highlighted the rapid growth in AI-generated child sexual abuse material and the need for Governments and technology companies to act urgently.