Is It Still You? Attributing Authorship and Authenticity in AI-Assisted Romantic Communication

要旨

In intimate messaging, how a difficult note is produced signals effort and ownership. We study how AI assistance level (light tone rewrite vs heavy full draft) and a brief co-sign disclosure shape receiver attributions and outcomes in apology and boundary setting. Study 1 (N=152) instrumented authoring to build a many-stimuli corpus. Study 2 (N=704) tested effects in a mixed effects experiment. Heavier drafting reliably reduced perceived ownership and authenticity; clarity/competence gains did not compensate. In apologies, co-signing a tone rewrite increased authenticity and forgiveness; co-signing a full draft decreased both. In boundary requests, co-signing yielded small or negative shifts. Stimulus-level analyses tied idiosyncratic "voice" cues to ownership/authenticity, and revealed sender-receiver miscalibration. We contribute: (i) scenario-aware causal estimates for help and disclosure; (ii) an empirically grounded, scenario-aware attributional account alongside a competence–integrity dissociation; (iii) evidence of sender–receiver miscalibration; and (iv) design guidance voice preserving defaults, ownership restoring scaffolds, and CPM disclosure.

著者
Guangrui Fan
Taiyuan University of Science and Technology, Taiyuan, Shanxi Province, China
Dandan Liu
Universiti Malaya, Kuala Lumpur, Malaysia
Lihu Pan
Taiyuan University of Science and Technology, Taiyuan, shanxi, China

会議: CHI 2026

ACM CHI Conference on Human Factors in Computing Systems

セッション: Romance and Relationships in the Age of AI

P1 - Room 118
7 件の発表
2026-04-16 18:00:00
2026-04-16 19:30:00