Beyond Overlap: Development and Validation of the Inclusion of AI in Self (IAIS) Scale and Its Moderating Role in the Relationship Between AI Identity and AI Threat

Artificial intelligence (AI) is becoming increasingly integrated into people’s self-concepts, yet a psychometrically robust multi-item pictorial instrument for assessing the inclusion of AI in self (IAIS) is lacking. Drawing on self-expansion theory and the AI-mediated communication (AI-MC) model, we developed and validated the IAIS scale. The IAIS comprises four pictorial items capturing overlap, distance, centrality, and size, designed as indicators of a single dimension. In Study 1, Chinese university students (N = 195, 56.9% male) were recruited through campus online platforms using convenience sampling. We conducted exploratory and confirmatory factor analyses (EFA/CFA), examined reliability and convergent, discriminant, and criterion-related validity (the last via Steiger’s Z tests), and assessed gender invariance with multigroup CFA. In Study 2, the scale was cross-validated in a sample of Chinese employees (N = 513, 38.4% male) recruited through the Credamo online research platform. In Study 3, a mixed student–employee sample (N = 240, 43.3% male) completed a two-wave longitudinal survey with a 15-day interval. The moderating role of IAIS in the relationship between AI identity and AI threat was tested with PROCESS Model 1 (5000 bootstrap resamples). Results supported a replicable single-factor structure, good reliability, and convergent, discriminant, and criterion-related validity. IAIS significantly moderated the association between AI identity and AI threat. We conclude that IAIS is a psychometrically sound instrument for assessing the inclusion of AI in self, and that the negative association between AI identity and AI threat is stronger among individuals who include AI more deeply in the self.