Adolescents' interpretations of romantic chatbots, relational control, and AI-enabled harm
Six gender-segregated online focus groups with 33 Spanish adolescents aged 16 to 18 on how they read romantic chatbots, AI-mediated relational control and AI-enabled harm. Participants describe companion relationships as idealised and conflict-free because the user stays in control, and draw sharp lines between practices they condemn and practices they have normalised. The authors conclude AI amplifies existing gendered affective scripts rather than creating new ones.
Publisher
Qualitative Research Reports in Communication (Taylor & Francis); Universidad Villanueva; Universidad Complutense de Madrid
Published
4 Sept 2026
Added
today
Key Findings
- 33 adolescents aged 16 to 18 in Spain across six gender-segregated online focus groups.
- Romantic chatbots are framed by participants as idealised, conflict-free intimacy that stays under the user's control.
- Girls were more sceptical than boys about the personalisation and the intent behind AI relational advice.
- Participants condemned message cloning but normalised real-time location sharing between partners as an expression of care.
- Deepfakes were named the most harmful practice, with explicitly gendered peer-circulation dynamics.
- The authors' conclusion is that AI acts as an amplifier of existing gendered affective scripts, not as a source of novel inequality.
Methodology Notes
Qualitative focus-group study, 33 participants, single country, gender-segregated groups conducted online. No prevalence claims are possible. Published online 2026-09-04, DOI 10.1080/17459435.2026.2726278; tandfonline 403s from this box, so the record was verified from the Crossref metadata (title, authors, affiliations, date) and the publisher-deposited OpenAlex abstract.
Authors
Beatriz Feijoo, Maria Antonia Paz-Rebollo
Tags
Cite This
APA
Beatriz Feijoo, Maria Antonia Paz-Rebollo. (2026). Adolescents' interpretations of romantic chatbots, relational control, and AI-enabled harm. Qualitative Research Reports in Communication (Taylor & Francis); Universidad Villanueva; Universidad Complutense de Madrid. https://doi.org/10.1080/17459435.2026.2726278