Revisiting associations between personality and attitudes towards AI by considering different AI-attitude-frameworks in Germany and Singapore

Mounting evidence suggests that personality is linked to attitudes towards AI, although empirical findings are heterogenous. Inconsistencies in the literature might have been caused by the application of different frameworks to assess individual differences in AI attitudes, but perhaps also by different ways to assess personality. Therefore, the present study investigated the Big Five of Personality (BFI-45) in the context of three different general AI-attitude-frameworks (two single item measures, ATAI-scale, ATTARI-12). For this research, n = 247 study participants were recruited in Germany, which also filled in a new scale assessing views on “personal enhancement by artificial intelligence” (PEAIS). A second sample consisting of n = 368 participants were recruited in Singapore to further elaborate on personality-AI-attitude-associations. Here, participants filled in the aforementioned AI-attitude measures, the BFI-10 (a short screening-tool for personality) and the AI-Interaction Positivity Scale. The AI-IPS assesses positive affect when interacting with AI systems.In the German sample, the AI attitude scales showed a robust overlap with each other. Of interest, neuroticism was linked with all general AI-attitude measures. In contrast, more specific scale dependent associations were observed for conscientiousness, agreeableness and extraversion. Interestingly, the Big Five of Personality were not at all linked to the PEAIS.In the Singaporean sample some of the associations between personality and AI-attitudes as observed in the German sample could be replicated, others not. The AI attitude measures robustly correlated with each other again.The present study shows that various combinations of AI-attitude and personality measures might result in different associations. These observations can help to better understand the inconsistencies appearing in the field and might assist researchers in selecting the most appropriate survey tool for their studies.