The constructive, overreliant, and irresponsible use of artificial intelligence tools in academia: Personality correlates and implications for academic integrity

The integration of AI into academia presents both opportunities and challenges for academic integrity. The literature review suggests that AI usage in academia may align with three general profiles—Constructive (e.g., using AI to enhance academic skills), Overreliant (e.g., heavy dependence on AI for academic tasks), and Irresponsible (e.g., using AI-generated content without verification)—each with significant implications for academic integrity. To expand our understanding of this subject, this study was conducted with two objectives: first, to introduce and validate the Academic AI Usage Inventory (AAUI), and second, to examine relationships between AI usage profiles and personality features, including maladaptive personality traits, psychopathic traits, attachment styles, and self-esteem. Data were collected online from 171 university students (aged 18–52; M = 29.19, SD = 7.70). Confirmatory factor analysis (CFA) supported the three-factor structure of the AAUI, which demonstrated acceptable internal consistency. Correlations and multiple regression analyses revealed that Disinhibition and Antagonism were positively linked to Irresponsible usage, with Disinhibition also associated with Overreliant usage. Conversely, Anankastia was positively related to Constructive usage and negatively to Irresponsible usage. Additionally, Callous-Unemotional (CU) and Impulsive-Irresponsible (II) traits were positively associated with Irresponsible usage, while II also correlated with Overreliant usage. Furthermore, Anxious attachment was associated with Overreliant and Irresponsible usage, while higher self-esteem predicted Constructive usage, and lower self-esteem was linked to Overreliant and Irresponsible profiles. Findings highlight the role of personality features in various AI engagement styles in academia, underscoring the need for policies to promote responsible AI usage and academic integrity. • Disinhibition and Antagonism linked to Irresponsible academic AI usage. • Anankastia positively linked to Constructive and negatively to Irresponsible AI use. • Psychopathic traits linked to Overreliant and Irresponsible academic AI behaviors. • Anxious attachment style predicts Overreliant and Irresponsible AI usage profiles. • Higher self-esteem fosters Constructive, while lower drives Irresponsible AI use.