AI is changing child exploitation. Our response needs to change just as fast.

As CameraForensics launches its Child Safety Online 2026 report, Dave Ranner reflects on the growing normalisation of AI-enabled abuse, the responsibility to build safer technology and why we cannot afford to wait for the perfect solution. With invaluable practitioner insights from Onemi-Global Solutions, the Internet Watch Foundation (IWF), Thorn and the Canadian Centre for Child Protection.

Technology has always presented both opportunities and risks. But the pace at which artificial intelligence is evolving is creating challenges that society, industry and those working to protect children are struggling to keep up with.

At CameraForensics, we believe technology has extraordinary potential to make the world safer for children. It's why we exist. But we also recognise that the same advances are being exploited to cause harm, often in ways that were difficult to anticipate even a few years ago.

Our latest Child Safety Online report explores how AI is changing the scale and scope of online child sexual abuse and exploitation, bringing together research and practitioner insights from organisations at the forefront of child protection.

For me, one of the most concerning findings isn't simply how sophisticated these technologies have become. It's how accessible, commercialised and increasingly normalised some forms of abuse now are.

When exploitation becomes normalised

For many years, discussions about online child exploitation have understandably focused on offenders operating in the darkest corners of the internet. Those threats remain. But AI is changing the landscape.

Tools capable of generating abusive imagery are no longer confined to technically sophisticated offenders or hidden online communities. They're increasingly accessible through mainstream digital services, with commercial ecosystems helping to make them available.

We're also seeing children use these technologies to create abusive images of other children, and that's something we need to confront. Even the language we use can minimise what's happening. Terms such as 'nudifying' or 'undressing' apps make the technology sound almost trivial.

But we're talking about technology being used to create sexually abusive images of real people, including children. The consequences are anything but trivial.

Our report highlights the scale of this emerging threat:

• 125% increase in sexually explicit deepfake images and videos of children reported to the Canadian Centre for Child Protection’s Cybertip.ca between 2022 and 2025.
• More than a 260-fold increase in photorealistic AI-generated child sexual abuse videos identified by Internet Watch Foundation analysts in 2025 compared with 2024.
• 19% of confirmed reports of child sexual abuse material sent to the UK’s Report Remove helpline between January and September 2025 involved manipulated imagery, including material created using AI and nudification apps.

These figures are deeply concerning. But behind every statistic is the potential for real harm to a child, and the risks extend beyond imagery.

Our report also explores financially motivated sexual extortion, sadistic online exploitation, grooming and the influence of AI chatbots and companions on young people.

The concern isn't only what these technologies can generate. It's how interacting with them might normalise harmful attitudes and behaviours, including misogyny, sexual exploitation and the dehumanisation of others.

As a parent, as well as someone working in this field, I find that particularly troubling.

The technology may be artificial. The harm is real.

One of the most complex challenges facing investigators is distinguishing AI-generated imagery from camera-captured material.

It's an important technical distinction, but one that can lead to a dangerous misunderstanding. We cannot simply categorise one image as 'real' and another as 'fake' and assume that only the former matters.

An AI-generated image depicting a real child can cause significant psychological harm. It can also indicate that a child is at risk, particularly when an offender has access to that child.

The same applies to sexual extortion. The fear, shame and distress experienced by a victim are no less real because the material used to threaten them was generated or manipulated using AI.

For investigators, this creates new challenges in identifying victims, assessing risk and prioritising resources. As the volume and sophistication of AI-generated material increase, so does the pressure on already stretched investigative teams.

Our response must recognise that the technology may be artificial, but the harm it causes is very real.

Safety cannot be an afterthought

I am not someone who believes we should stop developing technologies that have the potential to benefit society, quite the opposite.

But we need to become far better at anticipating how those technologies might be misused, particularly when the consequences affect vulnerable people.

Too often, we develop and release technology, discover the harm it can cause and then try to introduce safeguards afterwards.

We need to reverse that thinking.

Safety by Design means considering potential risks from the earliest stages of technology development, rather than treating safeguarding as something to address once a product has reached the market. That responsibility extends across the entire technology ecosystem, from AI developers and platform providers to the companies that distribute and monetise their services.

Regulation also has a critical role to play. Industry has had considerable opportunity to demonstrate that self-regulation can adequately protect children and has failed. The evidence increasingly suggests that stronger accountability and clearer legal responsibilities are needed.

The solution requires a collective effort, but responsibility is not shared equally.

Technology companies and platform providers must do more to prevent their products and services from enabling abuse, while policymakers need to establish and enforce stronger safeguards.

At the same time, researchers, law enforcement, child protection organisations and technology innovators need the support, resources and collaboration to identify emerging threats, develop effective solutions and protect more children.

We all have a part to play, but those with the greatest power to prevent harm must take greater responsibility for doing so.

We need shared responsibility, meaningful collaboration and a willingness to act.

Because if we don't change how we approach safety today, we risk having exactly the same conversation in five or ten year’s time about a completely different generation of technologies.

Turning technology into a force for good

While the challenges are significant, there are also reasons to be optimistic.

AI is creating opportunities to strengthen investigations, uncover hidden intelligence and help those working on the frontline identify and protect more children.

At CameraForensics, our focus is on translating advanced research and technology into practical investigative capabilities.

This includes developing AI image detection capabilities to help investigators understand whether material has been generated or manipulated, alongside uncovering forensic signals that can provide valuable intelligence.

But technology should support investigative expertise, not replace it. The objective is to reduce time-consuming processes, surface connections that might otherwise remain hidden and give investigators more capacity to apply their experience and judgement where it matters most.

And we cannot achieve that in isolation.

The insights shared by Onemi-Global Solutions, the Internet Watch Foundation, Thorn and the Canadian Centre for Child Protection have been invaluable in shaping this report.

Each organisation brings a different perspective, different expertise and a different part of the solution.

We see that same principle in our own work, collaborating with law enforcement, NGOs, academics and technical specialists to understand emerging threats and develop technology around real investigative needs.

No single organisation has all the answers. But by bringing the right people and capabilities together, we can make meaningful progress.

We cannot wait for the perfect solution

It's easy to look at the scale of these challenges and feel overwhelmed. But I'm continually inspired by the extraordinary people and organisations dedicating their expertise to protecting children.

Their work is a reminder that progress is possible.

One of my strongest beliefs is that we cannot keep waiting for the perfect solution before taking action. There will always be reasons why an approach might not work, why more research is needed or why the challenges are too complex.

Of course, we need evidence, rigour and appropriate safeguards. But we also need the willingness to test ideas, learn, collaborate and move forward. Because while we're debating the best response, technology continues to evolve. And children continue to be harmed.

We need to act with the same urgency and ambition as those exploiting these technologies for harm.

That's why we've produced the Child Safety Online 2026 Report.

Not because we have all the answers, but because we believe bringing together the evidence, expertise and perspectives of those working on these challenges is an important step towards finding them.

Our hope is that the report helps inform better decisions, encourages more meaningful collaboration and strengthens the collective response to AI-enabled child exploitation.

The future of technology is still being shaped. We have a responsibility to make sure children's safety is part of that future from the very beginning.

Download Child Safety Online 2026 Report

How is AI changing the scale and scope of online child exploitation?

Explore the latest research, emerging threats, investigative challenges and opportunities to build a safer digital future for children.

DOWNLOAD THE REPORT