IMPACTeen: Intentions, Manipulation, Persuasion, Annotations, and Consequences in Teen Communication Dataset
arXiv:2606.16910v1 Announce Type: new Abstract: IMPACTeen is a dataset of textual social influence scenarios spanning interpersonal, media-based, and digital settings in an adolescent context. It contains 1,021 texts, 5,100 individual annotation records, and gold labels for social influence techniques, with each text annotated from five distinct perspectives: teenagers, parents, psychologists, communication experts, and teachers. The resource was constructed through constrained LLM generation, followed by a two-step human editing and validation phase aimed at ensuring youth-context realism. A
The increasing sophistication of LLMs and the growing concern about their impact on youth underscore the immediate need for datasets to study social influence.
This dataset provides a critical resource for understanding, detecting, and potentially mitigating harmful social influence techniques, especially in vulnerable adolescent populations.
Researchers and developers now have a categorized, multi-perspective dataset specifically designed to train AI models to recognize and analyze manipulation, persuasion, and other social influence tactics in teen communication.
- · AI researchers
- · Social scientists
- · Educational technologists
- · Parents and educators
- · Malicious actors using social influence
- · Unregulated AI platforms if abuse is identified
AI models trained on IMPACTeen could become better at identifying social influence in online interactions.
Social media platforms might deploy such AI to flag or counter manipulative content targeting teenagers, potentially increasing platform safety.
This could lead to new educational tools designed to teach media literacy and critical thinking skills to adolescents, empowering them to resist manipulative tactics.
This signal links to a primary source. Continuum Brief monitors and indexes it as part of the live intelligence stream — we do not republish source content.
Read at arXiv cs.CL