General Attitudes towards Artificial Intelligence Scale (GAAIS): Hungarian adaptation and links to personality traits.

Background: The present study undertook the adaptation and psychometric validation of the Hungarian version of the General Attitudes toward Artificial Intelligence Scale (GAAIS) to assess both positive and negative attitudes toward artificial intelligence (AI) in relation to psychosocial functioning and personality traits. Methods: The adaptation followed international test-adaptation standards, involving translation, back-translation, and expert review. A total of 704 participants (557 women, 144 men) aged 18-60 years (M = 27.8, SD = 10.6) completed the GAAIS together with several validated self-report measures: the Mental Health Continuum-Short Form (MHC-SF), Self-Concept Clarity Scale (SCCS), frequency of AI usage, Problematic Internet Use Questionnaire (PIUQ), and Schizotypal Personality Questionnaire-Brief Revisited (SPQ-BR). Results: The Hungarian version showed solid internal consistency (Cronbach's α = 0.85 for the positive and 0.81 for the negative subscale) and a clear two-factor structure, supported by confirmatory factor analysis (CFI = 0.951, RMSEA = 0.058). The frequency of AI use in daily life emerged as the strongest predictor of both positive and negative attitude scores lending further support to the construct validity of the scale. The association analysis revealed that the behavioral components of AI-related attitudes are shaped by the competing motivational forces-approach (positive) and avoidance (negative). Specifically, the frequent use of AI is linked to the positive attitudes of GAAIS. In contrast, the unfavorable use of AI is associated with the negative attitudes of GAAIS. In the affective domain, anxiety sensitivity is associated with a negative attitude, and in the cognitive domain, schizotypal cognitive characteristics and difficulties in self-integration are linked to elevated negative attitudes in GAAIS. However, on the other pole of this cognitive dimension, adequate self-integration does not play a significant role in the formation of an AI-related positive attitude. Conclusion: These findings confirm the reliability and validity of the Hungarian GAAIS and highlight the importance of personality traits in shaping adaptive and maladaptive attitudes toward AI. The results underscore the value of a multidimensional framework for understanding AI attitudes. Adaptive traits were associated with psychological resilience, effective self-regulation, and constructive digital engagement, whereas maladaptive traits were correlated with social anxiety and problematic interactions with the internet and artificial intelligence (AI) technologies. A critical question remains: What outcomes may arise from when individuals hold positive attitudes toward AI but simultaneously experience difficulties with self-integration? This paradox highlights the need for further research into the complex interplay between personality structure and digital adaptation.