The online safety of children has reached a new level of concern, and this time the warning came from UK authorities carrying real institutional weight.
The NCA (National Crime Agency) and the IWF (Internet Watch Foundation) published a landmark guide aimed at parents and caregivers, with a recommendation that might sound extreme at first glance: don’t post photos of children on public social media profiles.
The reason goes way beyond what most people realize. Artificial intelligence tools available to the general public are being used by criminals to turn ordinary photos, like ones from a birthday party or the first day of school, into child sexual abuse material. And the most alarming part of all this? Most parents simply have no idea this is happening.
In 2025, the IWF identified 8,029 realistic images and videos of AI-generated child sexual abuse material, a 14% increase from the previous year. This isn’t some distant or futuristic scenario. It’s already reality, and it’s happening right now. 👇
How AI Is Being Used to Generate Images of Children
For a long time, the conversation around artificial intelligence focused on the benefits it could bring: automation, medicine, education, productivity. But there’s a dark side growing in parallel, far from the spotlight, and it directly involves the online safety of children. AI image generation tools, which anyone can access with just a few clicks, are being adapted and exploited by criminals to create child sexual abuse content from completely ordinary photos pulled straight from public social media profiles.
According to Lorna Sinclair, child sexual abuse education manager at the NCA, the average parent or caregiver doesn’t post a photo thinking it could be copied and turned into CSAM (child sexual abuse material). Many parents and caregivers don’t even know this problem exists. And it’s precisely that lack of awareness that makes the situation so dangerous.
The process is technically called image manipulation through generative models. In simple terms, it means an innocent photo of a child playing in the park can be processed by an AI model and turned into something completely different, without any physical crime needing to take place for the material to be produced. This represents a fundamental shift in the nature of the problem, because it eliminates barriers that previously existed.
Receive the best innovation content in your email.
All the news, tips, trends, and resources you're looking for, delivered to your inbox.
By subscribing to the newsletter, you agree to receive communications from Método Viral. We are committed to always protecting and respecting your privacy.
Before, criminals needed to make direct contact with victims, a process known as grooming. Now, with publicly available AI tools, that direct contact is no longer necessary. As Tim Wright, senior manager at the NCA, points out, all it takes is for the original image to be accessible on the internet for harm to be caused. This completely changes the scale and speed at which this type of crime can be committed.
Real Cases That Set Off the Alarm
The guide’s publication didn’t happen by accident. It came after a series of real cases that showed the true size of the threat. The IWF, which monitors CSAM incidents and maintains a reporting hotline, was contacted by minors who had been blackmailed by extortionists after their images were altered by AI to appear as nudity.
A confidential service called Report Remove, created to take down explicit images of minors shared without consent, also reported situations where regular selfies, with the person fully clothed, were converted into extreme pornography using artificial intelligence. In another case handled by the Childline service, a 15-year-old girl said a stranger had created a highly convincing fake nude of her using her face and her bedroom, apparently with material taken from her Instagram account.
There were also cases where UK school websites were targeted by blackmailers who copied photos of students, used AI tools to convert them into child sexual abuse material, and then threatened to publish the results. In response, a British advisory group on online harms, the EWWG (Early Warning Working Group), which includes the NCA and the IWF among its members, recommended that schools remove identifiable images of students’ faces from their websites and social media.
Why Public Profiles Are the Main Vulnerability
When a parent posts a photo of their child on Instagram, Facebook, or any other platform with a public setting, that image becomes available to anyone in the world, no exceptions. It doesn’t matter if the profile has a thousand or ten thousand followers, it doesn’t matter if the account belongs to an anonymous family or a digital influencer. If the profile is public, the photo is public, and that means it can be saved, copied, processed, and reused without anyone’s knowledge or consent.
Dan Sexton, CTO of the IWF, admitted he was very uncomfortable having to tell parents not to post photos of children, but said he saw no other option given the lack of existing protections. In his words, he would be very cautious about publishing images of children online, because right now there is no real protection against this kind of misuse. It’s a tough statement, coming from someone who deals with this problem firsthand.
The issue is made worse by the fact that social media platforms have been increasingly struggling to detect and remove this type of manipulated content. Moderation algorithms were trained to identify sexual abuse material based on known visual patterns, but AI-generated content frequently doesn’t follow those same patterns, slipping past automated filters with alarming ease. This creates a window of time during which the material circulates freely, causing damage that is virtually impossible to fully reverse.
What Parents and Caregivers Can Do Right Now
The guide makes it clear that the intention isn’t to tell parents how to behave online, but to raise awareness about the problem and about ways to tackle it. The core recommendation is to avoid posting images of children on public profiles, but that doesn’t mean giving up on sharing important moments from your kids’ lives. It means rethinking how that sharing is done and who it’s directed toward.
Authorities suggest three practical steps that can be taken today:
- Check the privacy settings on your social media accounts, making them private whenever possible;
- Review who can see images of your children by creating a close friends list or limiting visibility to selected people;
- Have open and calm conversations about the permission given to people and organizations to post images online.
The guide also recommends doing a kind of audit on your own accounts. It’s worth checking whether the child’s face, body, or school uniform can be seen in older photos, evaluating whether you still feel comfortable with that image being online, and seeing if it can be deleted or made private. Many of these photos were posted years ago, before AI advances made image manipulation this accessible, and they deserve a fresh look.
Another point raised is the need to talk to other family members and close friends. In many cases, grandparents, aunts, uncles, and godparents also post photos of the children without realizing the risks involved. A child’s online safety depends on the behavior of the entire social circle around them, not just the parents. It’s worth having a clear and relaxed conversation about what steps can be taken regarding these posts, including older ones.
Reviewing Consent and the Right to Removal
A very important aspect of the guide is the recommendation to review consent forms signed at schools, daycares, and sports clubs. Many of these documents were signed before AI advances, when image manipulation wasn’t such an obvious concern. Authorities suggest that parents reassess these agreements and consider withdrawing permission if they no longer feel comfortable with it.
Tom Dyson, head of marketing at the IWF, emphasizes that parents have the right to request the removal of their children’s photos. According to him, if you want a photograph of your children taken down from a website or social media platform, you have every right to ask for that. It’s simple information, but something a lot of people don’t know, and it can make a real difference when it comes to protecting a child.
The videos released as part of the guide show fictional scenarios of parents photographing their children in everyday situations, like playing sports or at the school gate, and being reminded of the risks of sharing those photos online. The goal is to encourage parents and children to say no to sharing whenever they feel uncomfortable with it.
The Conversation That Still Needs to Happen
What makes this situation even more delicate is the knowledge gap between what’s happening technically and what most parents know about the subject. Artificial intelligence is still a topic many people associate with robots, science fiction movies, or complex corporate tools. The idea that an AI model can process a photo posted on social media and turn it into sexual abuse material sounds surreal to anyone who doesn’t closely follow tech discussions. But that is exactly the reality UK authorities are flagging with urgency.
Digital education needs to stop being a secondary topic and start taking center stage in conversations at home, in schools, and in public policy. This isn’t about creating panic or turning the internet into a place families should avoid, but about understanding that the digital landscape has changed dramatically in recent years. Sharing habits that were considered safe five or ten years ago need to be reassessed in light of artificial intelligence’s new capabilities.
It’s worth noting that organizations like the NSPCC also recommend that minors keep their social media accounts set to private, reinforcing that protection needs to start early. In the United States, the National Center for Missing & Exploited Children (NCMEC) plays a similar role to the IWF, receiving reports and working with authorities to remove illegal content from the internet. Knowing about these channels is part of a more informed approach to the digital risks involving images of children. 🛡️
The warning coming from the UK is not an isolated case. Child protection organizations in many countries are monitoring the same phenomenon and reaching the same conclusions. The speed at which AI tools are evolving is far greater than the speed at which laws, platforms, and public awareness can keep up. That’s why, for now, the main line of defense is still information and the choices each family makes every day about how they share images of children in the digital world. 🔍
