Social Work Students’ Views Toward AI and Ethics: The Role of Technical and Ethical AI Literacy

Background and Purpose: As artificial intelligence (AI) use rapidly increases within social work education and practice settings, the profession faces a critical need to understand how future practitioners grapple with the ethics of AI integration. Drawing on the Theory of Planned Behavior (TPB), this study explored the factors influencing AI ethical attitudes and adoption among social work students, with particular attention to how self-perceived technical and ethical AI knowledge shape attitudes toward AI’s ethical use in both educational and professional practice settings.

Methods: A cross-sectional survey was administered to social work students (N = 97) at a public university in the Southeastern United States, including participants from both undergraduate and graduate levels. The survey measured self-perceived technical knowledge of AI, self-perceived ethical knowledge regarding AI, attitudes toward the ethical integration of AI in education and professional practice, and current AI use behaviors. Data were analyzed using exploratory descriptive and subgroup analyses followed by a series of linear regression models to assess preliminary associations among perceived knowledge, ethical attitudes, and reported use.

Results: Analyses revealed several pathways connecting technical and ethical literacy to AI-related behaviors. Greater self-perceived technical knowledge was a significant predictor of greater perceived AI ethical knowledge (b = .43, p < .001). In turn, greater perceived ethical knowledge was associated with less positive views toward AI use in educational settings (b = -.40, p < .001), but more positive views toward AI use in social work practice (b = .23, p < .05). Practice-related attitudes were also higher among older students and those enrolled in practice-oriented part-time programs (b = .61, p < .05). Finally, more positive ethical views in both educational and practice settings were associated with greater current AI use (b = .22, p < .05; b = .29, p < .05, respectively).

Conclusions and Implications: These exploratory findings suggest that technical literacy often acquired through hands-on engagement with AI tools may be a foundational factor shaping emerging social workers’ ethical discernment. The contrast between student skepticism about AI in education and relative optimism about its use in practice may reflect nuanced concerns about academic integrity alongside recognition of AI’s potential to address longstanding service delivery challenges. These findings underscore that technical training must not be siloed from ethical discourse; rather, technical proficiency appears to facilitate the critical thinking required for ethical AI integration. Future research should prioritize more generalizable samples, longitudinal designs, and the development and evaluation of educational modules that cultivate a literate, ethically grounded AI foundation for the next generation of social workers.