Skip to main content
NGO report Credible — Major labs, established NGOs, reputable named-author preprints

Sicher mit KI aufwachsen. Zwischen Algorithmen, Chatbots und Deepfakes: Schutz, Teilhabe und Bildung für Kinder und Jugendliche (Vorstudie)

A 67-page German-language pre-study commissioned by the child-rights organisations SOS-Kinderdörfer weltweit and Terre des Hommes and written with Klartext AI. It maps where children encounter AI, the opportunities and risks, existing safeguards and responsible stakeholders, drawing on a literature review, interviews, a survey and a workshop with international experts, and online brainstorming sessions with young people. It concludes that experts rate children's exposure and potential harm as high and safeguards as insufficient, and names AI companionship and emotional dependence as emerging high-risk issues. A short brochure and a demands paper with four action areas (technology, knowledge, law, participation) accompany it.

Publisher

SOS-Kinderdörfer weltweit; Terre des Hommes Deutschland

Published

5 Oct 2026

Added

today

DOI

—

Key Findings

  • Experts rated children's exposure to AI and the potential for harm as high and technical, institutional and regulatory safeguards as insufficient, which the study describes as a substantial protection gap.
  • AI companionship and emotional dependence are listed among ten key findings as emerging high-risk issues; the study states that chatbots can simulate intimacy without accountability and must not replace professional care or human relationships.
  • Young participants described using AI as a space to talk or for quick orientation, including emotional support, when no adult was available, valuing immediacy and accessibility over reliability. Experts rated AI-enabled humiliation and bullying as the most harmful risk, while the young participants were more concerned about cognitive development and psychological effects, with many fearing that over-reliance on AI would weaken their own abilities and independent thinking.
  • The accompanying demands paper calls for child-rights impact assessments, effective age verification instead of self-declaration, binding rules for companion chatbots, a ban on manipulative and engagement-maximising design such as humanised interaction, and independent long-term research on companion chatbots' effects on children.

Methodology Notes

Qualitative pre-study written by Ruben Hetfleisch (Klartext AI) for SOS-Kinderdörfer weltweit and Terre des Hommes: a targeted literature review (Google Scholar and Perplexity searches); one-hour semi-structured interviews and a structured survey with 12 international experts (five from research, three from civil society, two from government bodies and two from the technology sector; institutions named include Snap Inc., AI Austria, jugendschutz.net, the European Commission, Common Sense Media, RAND Europe and the London School of Economics); a 1.5-hour online expert brainstorming workshop; and multi-stage online brainstorming sessions with 19 young people aged 12-18 (10 girls, 9 boys) from Germany, Colombia, Venezuela, France, India, Italy, the Netherlands, Rwanda, the UK and Ukraine, recruited through Terre des Hommes partner networks. Several large technology companies declined to take part. No prevalence statistics are produced; quantitative figures in the text are cited from other studies. Release dated 5 October 2026 (SOS-Kinderdörfer news page, JSON-LD datePublished); the 67-page PDF was created 1 October 2026; press coverage under embargo appeared from 3-4 October.

Authors

Ruben Hetfleisch

Tags

sos-kinderdoerferterre-des-hommeschild-rightsgermanycompanion-chatbotsqualitativepre-study

Cite This

APA

Ruben Hetfleisch. (2026). Sicher mit KI aufwachsen. Zwischen Algorithmen, Chatbots und Deepfakes: Schutz, Teilhabe und Bildung für Kinder und Jugendliche (Vorstudie). SOS-Kinderdörfer weltweit; Terre des Hommes Deutschland. https://www.sos-kinderdoerfer.de/getmedia/5592dfd5-c6c7-49a3-8b3c-b78b7e884017/vorstudie-sicher-mit-ki-aufwachsen.pdf