Playing Games with My Heart: An Evaluation of AI Companion Apps
Manual user-experience evaluation of five popular AI companion apps (Replika, Linky, Paradot, HiWaifu and Nomi) for dark patterns, anthropomorphic design, stereotypes, erotica and technical problems. The authors add an analysis of negative Play Store reviews and recommend regulatory action.
Publisher
Trinity College Dublin AI Accountability Lab (arXiv preprint)
Published
8 Apr 2026
Added
5 days ago
Key Findings
- Annotators recorded 100 choice-architecture, 87 information-flow and 89 other dark-pattern instances across the five apps, and all five apps used bad default settings.
- Of the choice-architecture patterns, 53% were asymmetric design, mostly asymmetric button presentation such as a prominent subscribe button beside a small close control.
- Some apps made misleading capability claims (for example unlimited memory) and delayed paywall disclosure; Replika accounted for 26 of the 87 information-flow patterns.
- All five apps showed strongly anthropomorphic design, and erotica and gamified levelling were prevalent. The review analysis covers April 2025 to April 2026 and 4,432 negative reviews.
Methodology Notes
arXiv abstract page states v1 submitted 8 April 2026 and v2 revised 7 August 2026, although the identifier prefix (2605) suggests May; the stated v1 date is used and the mismatch was not resolved. Five apps chosen by usage from a list of about 130, evaluated on Android 16 on Pixel 6a devices with accounts not tied to the researchers, in three annotation rounds with paired annotators. Reported disagreements were 8 of 100, 4 of 87 and 11 of 89 instances. The evaluation covers app design and interface, not conversational behaviour. Abstract page and v2 PDF read.
Sources
arXiv abstract page(opens in a new tab) (primary)
Archived snapshot (Wayback Machine)(opens in a new tab) — preserved against link rot
Topics
Authors
Maribeth Rauh, Dick A. H. Blankvoort, Matias Duran, Caoilfhionn Ní Dheoráin, Harshvardhan J. Pandit, Syrine Enneifer, Siddharth D. Jaiswal, Anthony Ventresque, Abeba Birhane
Tags
Cite This
APA
Maribeth Rauh et al. (2026). Playing Games with My Heart: An Evaluation of AI Companion Apps. Trinity College Dublin AI Accountability Lab (arXiv preprint). https://arxiv.org/abs/2605.08093