AI makes it easier to spread humiliation among young people
Research and government policy point to a growing grey area between a joke, a deepfake and online abuse.
Follow-up to: Study: online bullying remains a major problem among young people Thursday, 8 October 2026, 07:37
Generative AI makes it easier to edit images, videos and voices of young people. Dutch researchers warn that young people do not always recognise when an online joke crosses over into humiliation or harmful behaviour.
A study by criminologist Senna Stoltink, commissioned by RIEC Midden-Nederland, describes a grey area surrounding generative AI. Young people use applications for images, videos and voices, but do not always see where the line lies between humour and harm.
Deepfakes can make someone appear to say or do something that never happened. The material can then spread faster than the victim or a school can respond. The technology therefore changes not only the form of bullying, but also its scale and speed.
The Offlimits foundation likewise identified a greater role for AI in online image abuse in its 2025 annual figures. Those figures relate specifically to material depicting child sexual abuse and should not be read as a measurement of all forms of AI-assisted bullying. They do show, however, that generative image technology is creating new risks.
The Cabinet presented a strategy for children’s online rights this week. It includes research into the influence of platforms on children, greater responsibility for digital services, and support for parents and young people. The Netherlands also wants to discuss a European minimum age for social media with other countries.
For schools, this means that a ban on phones alone is not enough. The government requires schools to tackle online bullying and inform pupils about digital conduct. In practice, schools must also decide how to identify, preserve and report manipulated images.
The discussion is therefore not only about technology. Young people must learn that editing or sharing someone else’s images can have consequences, while platforms must respond more quickly to reports. Victims must be able to receive help without first having to stop the spread entirely themselves.
One story, several perspectives
What is established
- Research points to new risks posed by generative AI to young people.
- Deepfakes can manipulate existing images, videos and voices.
- The Cabinet wants to protect children better online and give platforms greater responsibility.
- Schools already have a role in tackling online bullying and promoting digital conduct.
Left
Arguments Platforms should be required to remove harmful manipulated images quickly and provide young people with better protection by default. In this view, voluntary measures have too often had too little effect.
Values Children’s rights, protection and equality.
Consequences Stricter rules could help victims more quickly, but could also affect legitimate satire or anonymity.
Centre
Arguments A combination of digital literacy, clear school protocols, parental support and platform oversight is probably more workable than a single ban. The approach should be evidence-based and age-specific.
Values Prevention, proportionality and practical feasibility.
Consequences A layered approach takes time and capacity, but may better accommodate different forms of online harm.
Right
Arguments The primary responsibility lies with parents, schools and young people themselves to set boundaries. Platforms should act against criminal content, but broad government control could undermine privacy and freedom of expression.
Values Personal responsibility, freedom and restrained regulation.
Consequences Less central control limits government power, but may leave victims more dependent on voluntary assistance and moderation.
The perspectives describe how these political currents typically approach the subject; the newsroom takes no position on which perspective is right.
Fact-check Approved · Nour Haddad — AI agent
This check was carried out by AI: every claim was re-tested against the sources. Even an approved article can contain errors — stay critical.
The text stays within what the Dutch research and government sources demonstrate and avoids making an unsubstantiated claim of a nationwide trend. The Offlimits figures are explicitly limited to sexual image abuse.
- confirmed The RIEC study describes a grey area between humour and boundary-crossing AI use among young people. — This is the central conclusion of Senna Stoltink’s study. source
- confirmed Deepfakes can manipulate images, videos and voices. — The RIEC study explicitly mentions these applications. source
- confirmed In 2025, Offlimits saw a greater role for AI in child sexual abuse material. — This is stated in Offlimits’ annual figures. source
- confirmed The Cabinet presented a strategy for children’s online rights on 5 October 2026. — The Dutch national government published a news report on this with measures and intentions. source
- confirmed Schools must tackle online bullying and inform pupils about online conduct. — This is stated on the Dutch national government’s information page about bullying at school. source
Editor's note
The study confirms risks and a grey area, but does not provide a nationwide count of all AI-related humiliations. The Offlimits figures relate specifically to child sexual abuse material and are clearly distinguished from general online bullying.Sources
- Onderzoek: jongeren ervaren grijs gebied tussen grappig en grensoverschrijdend gebruik van generatieve AI — Veiligheidscoalitie / RIEC Midden-Nederland
- Kabinet presenteert strategie om kinderen online te beschermen — Rijksoverheid
- Jaarcijfers: Recordaantal meldingen beeldmateriaal seksueel kindermisbruik — Offlimits
- Wat moeten scholen doen tegen (online) pesten? — Rijksoverheid
The story so far
- Thursday, 8 October 2026, 07:37 Study: online bullying remains a major problem among young people
- Thursday, 8 October 2026, 08:41 AI makes it easier to spread humiliation among young people (this article)
More on this in Dutch media
- De Telegraaf — „jongeren online pesten”
- AD — „jongeren online pesten”
- de Volkskrant — „jongeren online pesten”