Online child sexual abuse and the protection of minors, with Véronique Béchu

How has child sexual abuse evolved in recent years? Are there any “distinctive features” of “online child sexual abuse” (compared to offline child sexual abuse)?

It is difficult to distinguish between online child sexual abuse and offline child sexual abuse. There is a true continuum of abuse, as acts committed online are often linked to physical acts in the real world. Even when minors are victims of online sexual extortion (sextortion), the ultimate goal is frequently to arrange in-person meetings. In some cases, the perpetrator belongs to the minor’s circle of trust (which consists of family members, close friends, and people whose roles allow them to exercise authority over the minor).

A recent example of a relatively common pattern of predation: the indictment of a middle school supervisor for rape after he “groomed”¹ minors at his school and then threatened them to coerce them into meeting him in person. When it comes to sexual violence against children, the Internet is just one of many means used to achieve this end.

It’s also important to remember that, legally speaking, viewing, possessing, distributing, or importing child sexual abuse material is already a criminal offense in itself.² Yet perpetrators often defend themselves by saying they “didn’t touch” the child, but they are indeed participating in a system in which a victim is assaulted, filmed, photographed, and then “revictimized” every time the content is viewed or shared. For content to be consumed, it must first have been produced. There is therefore necessarily a victim behind every image or video.

As for members of the online child sexual exploitation “community,” they most often operate in isolation. These individuals seek, above all, anonymity. Exchanges are generally based on a logic of reciprocity, grounded in the value attributed to the content due to its exclusivity, the very young age of the victims, or its rarity. The goal is often to expand a personal collection. While monetization does occur—particularly via the dark web, encrypted apps, or certain marketplaces—it remains relatively marginal. The exception is livestreaming, a phenomenon that originated in the Philippines—first within criminal networks exploiting street children, and later among very poor families who saw it as a means of survival. These criminal acts are often committed by women or within family settings and are part of organized criminal networks.

Finally, even though the idea of a progression in behavior has not been definitively proven scientifically, real-world experience shows that pedophilic practices can follow a pattern of addiction and radicalization. Some individuals first stumble upon this content without seeking it out, then return to it to verify what they felt—which already constitutes habitual viewing. Next comes possession, then sharing, and finally the search for increasingly rare and violent content. This obsessive search often leads to the production of new content, either involving children from the perpetrator’s immediate circle or by contacting minors through online games, apps, or social media to obtain previously unseen content, images, or videos, and then engaging in sexual extortion to produce exactly the desired content. This escalation can lead to content marked by extreme violence, including acts of torture and barbarism, as well as a decrease in the age of the victims.

How is AI redefining online child sexual abuse?

Artificial intelligence has numerous implications for child sexual exploitation. In particular, it enables the large-scale production of content and the generation of images or videos that cater to the worst fantasies. Today, with certain tools available to the general public, sometimes all it takes is a simple prompt to obtain an image or a video. This opens up a vast field of opportunity for child sexual exploiters.

AI also facilitates the misuse of content that appears harmless at first glance. Photographs of children posted online can be altered or sexualized, then used for blackmail or extortion. This situation is exacerbated by the large number of images that some parents post on social media, a phenomenon known as Sharenting.³ In fact, a significant portion of the content circulating in these spaces originates from social media and posts by the families themselves.

Furthermore, AI is contributing to the emergence of other phenomena that could be harmful to minors. These include, in particular, AI companions, which currently constitute a largely unregulated field where the risks remain poorly understood.

Furthermore, they can lead to genuine addiction. A study by the Observatoire e-Enfance / 3018 shows that 87% of the young people surveyed use AI, and that among those with more extensive interactions, 10.6% report being addicted. The risks related to child sexual exploitation embedded in these agents are considerable; they can expose users to situations of increased vulnerability.

Several countries are currently experimenting with different approaches to protecting minors online: Australia has banned children under 16 from accessing social media, Brazil has passed a law requiring that accounts held by minors under 16 be linked to a responsible adult, and France is considering banning social media for children under 15. What are your thoughts on this?

Regulators often focus on the wrong problem—it’s like not seeing the forest for the trees. These measures are easily circumvented, even in strict regulatory environments like Australia. Above all, we must compel platforms to design tools that are truly safe for minors by incorporating protection requirements from the outset, in line with the “safety by design” approach.

A ban alone is not a sustainable solution. If social media is banned, the risks will simply shift elsewhere—particularly to gaming platforms, messaging apps, or other digital environments—on an even larger scale than is currently the case. Social media platforms are sometimes categorized based on their level of risk, but the dangers can be just as significant on any platform.

Is it okay to continue posting photos or videos of your children on social media? If “yes,” what are the best practices for adults, particularly parents (private accounts, blurring faces, etc.)?

It is important to reflect on the reasons behind these practices and to remind parents that children are not simply an extension of themselves. They are individuals in their own right, with rights of their own, including the right to one’s own image and the right to privacy. Furthermore, once they reach a certain age, they are capable of deciding whether or not to give their consent to such a publication. After this reflection, when sharing is deemed necessary—for example, to show photos to friends or family—it is essential to put strict safeguards in place, such as limiting the recipients and using restricted groups.

However, in most cases, sharing images is unnecessary, and because parents set their social media accounts incorrectly, they share photos or videos without realizing that they may be accessible to people with malicious intent—often due to a lack of awareness of the risks. Even when accounts are labeled “private,” accepting a large number of contacts significantly reduces that privacy. If a photograph is shared, it should remain within a limited and controlled circle. Furthermore, blurring faces is not sufficient protection. In most cases, what interests malicious individuals is not only the visible identity but also the pose or context of the image. And with an AI tool, once again, this measure—intended as a safeguard—can easily be removed. It is also essential to protect children by systematically asking for their consent before any content is posted.

Finally, what advice would you give to minors to help them use social media and digital tools in general as safely as possible?

It is not advisable for children to have access to the Internet too early. At age eight, for example, there is not necessarily any benefit to being exposed to these activities on their own. Yet this happens extremely often. The main issue here is digital guidance—that is, learning to use these tools with an adult mentor. Often, when parents give their children a phone—which is a significant milestone—they fail to provide any guidance on how to use it safely. The “if something happens or if something happens to you online, I’ll take your phone away” approach is completely counterproductive. This is because the child will hide any difficulties they face in order to keep the phone. It’s essential to spend time with the child, explain best practices, and set clear boundaries that include screen time limits, warning keywords, disabling chat features, and a strict rule against contacting strangers—and to assure them that no matter what happens, you’ll be there to protect them.
Therefore, it is important to raise parents’ awareness of resources such as the 3018 app, which is designed for young victims and witnesses of cyberbullying and digital violence, as well as the risks they face online.

More broadly, it is important to recognize that children’s online lives are now just as important—and in some cases even more formative—than their offline lives. It is their space for socializing. Parents must accept this reality and take an active interest in what their children are doing on the Internet. They should be able to ask their children about their online lives just as they would about their lives at school.

¹Grooming (or child luring) is a manipulation technique used to lure a minor for sexual purposes.
²Article 227-23 du Code pénal.
³A portmanteau word combining “to share” and “parenting”.