Protection of minors: AI announces itself as a problem

Contents

One thing is clear, said Stefan Glaser, head of jugendschutz.net: Almost all age cohorts of adolescents use the internet in some way, with audiovisual content taking center stage. "When we talk about protection and participation, we must therefore assume that services are also used by those for whom they are not actually intended," said the head of the organization, which is a joint competence center of the federal and state governments, at the presentation of the annual report in Berlin.

This leads to a wide variety of problems that arise at the office. The long-established organization works with information from third parties, which it checks and then reports to the operators via its own direct channels. A total of 45,963 reports to Jugendschutz.net were evaluated by the staff. In 7645 cases, violations of legal provisions were identified and responded to – although one such 'case' can also involve many related media.

Two thirds of the cases concerned sexualized violence, 12 percent pornography and sex, 11 percent political extremism, 5 percent self-harming content and two percent cyberbullying. The organization contacted providers over 3,000 times, usually successfully, and cases were handed over to the law enforcement authorities more than 3,500 times.

Young offenders an increasing problem

The youth protection officers are also struggling with a problem: in a total of 3,582 cases of child and youth pornographic content, there is an increasing amount of content in which the perpetrators are still minors and have sent each other photos, for example, which have been covered by the relevant section of the Criminal Code 184b since 2021. The traffic light government actually wanted to have this regulation reformed long ago.

In addition, jugendschutz.net is increasingly finding content created via video chats with minors. There is a suspicion that the creators could commit their crimes almost anonymously via cybergrooming, the targeted fraudulent gaining of minors' trust and subsequent remote abuse.

Hamas attack and consequences in the children's room

The spread of extremism and hate content was another important topic in the reporting period, said Glaser –, particularly after the Hamas attack on Israel and the events that followed. It is particularly problematic that children and young people also use audiovisual platforms to keep up to date with world events.

According to Glaser, highly violent and emotional content, and in some cases false or discriminatory content, could "certainly have a radicalizing effect". In the case of hatred against Muslims, hatred against Jews and the disparagement of other groups, AI is now increasingly being used – and is becoming increasingly difficult to recognize as such. According to the head of jugendschutz.net, modified propaganda has an impressive effect on them. Telegram in particular is an alternative platform for extremists due to its inadequate content moderation, but the content also finds its way onto other platforms.

Despite relevant new legislation, the providers' reporting systems are still not efficient: While reports submitted by jugendschutz.net as a normal user often did not lead to success, the providers took action in almost all cases of reports by Jugendschuz.net – In the case of Instagram, the difference was around 60 percent.

Youth media protection politicians see age verification as key

For Rhineland-Palatinate's Youth Minister Katharina Binz (Grüne), this is part of a bigger picture: the platforms must fulfill their responsibility -- and the legal framework must be applied consistently. Marc-Jan Eumann, Director of the Rhineland-Palatinate Media Authority and Chairman of the Commission for the Protection of Minors in the Media (KJM), used the example of porn platforms to illustrate that this has not always been the case: Despite the obligation under the Interstate Treaty on the Protection of Minors in the Media and the DSA –, these platforms still – do not have effective age verification systems. Yet there have long been functioning and even data-saving implementations: "There are no more excuses for providers when it comes to reliable and easy-to-implement age verification systems," said Eumann in Berlin.

This also applies to other platforms that, due to the age structure of their user base, should actually be able to reliably find out which age group their users belong to –, for example, to make certain age-appropriate default settings for them, such as non-public profiles. In principle, the Digital Services Act also provides for this, but in practice this has rarely happened to date.

Family Minister hopes for DSA

The internet is an important information space for children and young people, often their only one –, but with many problems, said Lisa Paus (Greens). They have a right to safe and carefree participation in the digital world, said the Federal Minister for Family Affairs. The platforms had been obliged to act with the DSA and she expressly welcomed the initial procedures of the EU Commission as the supervisory authority. However, the enforcement structures are still being established, said the Minister for Family Affairs: "We now have the law, now the question is to what extent we are in a position to enforce the law."

At the same time, she hoped for progress in the negotiations on the CSAM regulation to combat content that depicts sexual abuse. As part of these consultations, the BMFSFJ has commissioned a concept for age verification systems that protect fundamental rights, which is to be presented this year. Paus explained that she was confident that the dossier could be completed soon.

Media literacy should also help

The ministers, the supervisory authority and jugendschutz.net are also concerned with the question of how media literacy can be taught to the often young users. Paus announced that the BMFSFJ would like to see cross-state, common minimum standards for media literacy teaching in schools and is working towards this.

Meanwhile, Marc-Jan Eumann made an appeal to parents for media literacy: "Don't put a single picture of your children online with their names and without pixels," he urged citizens, and not just regarding processing by AI systems. "We need a sense of disturbance when a picture online shows a child, and then in the worst case, even with a name."

(vbr)