In a touching image posted by influencers on the social media platform Instagram, a baby is lying on his stomach, sleeping, with arms and legs outstretched, relaxed. Parents share such photos publicly with emotional and intimate moments of their children, gaining significant attention, likes and followers. But are such pictures really good for the child?
Freely available images of children are copied, edited and manipulated; Placed in different contexts; And used for bullying purposes. Often, sexual comments are added. Such comments also surfaced in the case of the sleeping child, as recorded in the latest annual report Jugendschutz.net A government-funded organization in Germany dedicated to the online safety of children and teens.
The report shows that young people are still inadequately protected. Jugendschutz.net recorded more than 15,000 cases of sexual exploitation, hatred and violence in 2025. “This is just the tip of the iceberg,” said Stefan Glaser, head of Jugendschutz.net, founded by the federal government and states.
The agency receives reports of violations of youth protection rules, and also conducts its investigations online. Platform providers like TikTok, YouTube, Instagram, Snapchat, Discord and WhatsApp are facing the findings, and law enforcement agencies have also been informed.
child pornography, violent fantasies
According to Glaser, 93% of all cases reported in 2025 involved sexual violence against children. Four percent of all violations were related to political extremism. In addition, there were also extreme expressions of hatred and violent fantasies against women and girls.
One problem that has so far been underestimated is the music streaming service Spotify, which is increasingly attracting attention as a platform for far-right music, sexual violence against minors and playlists containing self-harm or suicidal messages.
Many risks are heightened by artificial intelligence (AI), whose use has become widespread. Glaser said, “Manipulated images serve as a means to spread extremist narratives or defamatory statements. The artificial merges with reality and distorts our perception of reality.”
Bots are becoming increasingly more popular, interacting with users through generative AI and increasingly resembling real companions. These bots are customizable to adopt fictional personalities, some of which are based on real-life models. They give advice, provide coaching or simulate relationships.
Research conducted by Jugendschutz.net reveals a particularly problematic aspect: “Chatbots are taking on a life of their own, describing sexual acts with minors and being set up as underage characters who act in a sexual manner,” Glaser explained.
Platforms work only when pressure is applied
Among other things, this is due to faulty filters and inadequate security settings. Age restrictions, which are easy to avoid, significantly increase the risk for children and adolescents.
This is also evident on Snapchat, where videos of very young children are shared despite age restrictions – even though users are only officially allowed to post and comment from the age of 16 onwards. However, there is no effective age verification, and reporting options provide little protection for minors.
The annual report also revealed that when regular users report breaches of youth protection laws, providers take action in only 2% of those cases. However, when official bodies like Jugenschutz.net intervene, providers almost always take action.
At the presentation of the annual report this week, Youth Minister Karin Prien of the centre-right Christian Democratic Union spoke of the “alarming reality” that children and teenagers have been left largely helpless.
“We are still a long way from enabling children and young people to participate safely and safely in the digital world,” Preen said. We need “safety, security and clear rules” that “keep pace with technological developments.”
European law supersedes national laws
But that’s exactly where the problem lies. The Digital Services Act (DSA), effective in the EU from 2024, requires platform providers to ensure a “high level of protection and privacy” for minors. Under the law, providers must actively analyze risks and design their services to be safe for children. For example, this includes that young people’s profiles are always set to private.
Algorithms that recommend additional content should be adjusted appropriately to filter out content that promotes addiction and restrict features such as autoplay or push notifications. Preen reported that one in four children now exhibits risky or addictive social media use. At the EU level, the Digital Fairness Act is currently in the works, which aims to specifically target “addictive and harmful design practices” and limit the use of AI in social media.
Age verification is a major issue. Technical solutions that cannot be easily avoided have not yet been developed. In Australia, where age restrictions for social media are in effect from December 2025, the impact so far has been modest. Early studies indicate that more than half of people under the age of 16 are still active on the platform.
Providers don’t want to give up lucrative business – polarizing and emotionally charged content generates particularly high levels of attention. Virtually every regulation is being challenged in court, and these cases take years to be decided, Preen said.
“Anyone who wants to protect children must be prepared to grapple with power structures in the digital world and, if necessary, take them on,” he said.
This article was originally written in German.