Understanding AI Use, Knowledge, Concerns, and Learning Needs Among Social Science Students: A Multi-University Descriptive Study

Background and Purpose: Artificial intelligence (AI) is rapidly reshaping higher education by changing how students write, search for information, organize ideas, and complete academic work. Emerging research suggests that students often view AI as useful for efficiency, learning support, and task completion, but their engagement is uneven and frequently accompanied by uncertainty about accuracy, ethics, and appropriate use. Despite growing literature on AI in higher education, relatively little research has focused on students in social science fields, where disciplinary values such as critical thinking, ethics, human relationships, and attention to structural inequities may shape both enthusiasm for and resistance to AI. As the first paper in this symposium, this study provides a cross-university descriptive portrait of AI use, self-reported knowledge, concerns, perceived benefits, and learning conditions among social science students, laying the foundation for the subsequent studies.

Method: Data came from a cross-sectional online survey of 396 students across three universities. The survey was distributed via Qualtrics to students in social work and allied social science programs at three U.S. universities between May 2025 and March 2026. Participants were recruited through department listservs using convenience sampling. The sample included 302 graduate and 94 undergraduate students and was predominantly female (80.3%), white (67.4%), with a mean age of 26.9 years (SD = 8.3). Social work students comprised the largest disciplinary group (59.1%), followed by public health (12.1%), criminal justice (10.6%), and psychology (7.6%). The survey assessed four domains: AI use frequency, self-rated AI knowledge, attitudes toward AI, and AI learning needs. Analyses included descriptive statistics with chi-square tests and independent samples t-tests.

Results: AI use was concentrated in academically adjacent tasks rather than daily use. The most common uses were personal or professional writing assistance (83.4%), academic writing assistance (78.1%), and academic topic searches (78.1%). Undergraduate students reported higher use than graduate students across several tasks and had significantly higher AI self-knowledge scores (t=-3.26, p<.01). Students expressed strong awareness of AI-related ethical concerns (86.8%) and limitations (79.3%), but lower confidence in using AI effectively (53.8%) and ethically (38.6%), suggesting a gap between ethical awareness and ethical competency. Concern levels were high overall, especially regarding misinformation from AI hallucination (65.7%), hindering student learning (64.9%), and undermining humanistic aspects of work (55.6%). Although students viewed AI positively for accessibility (91.9%), many were uneasy about its role in social sciences education; only 23.4% responded it would help prepare an effective workforce. Over 70% felt inadequately trained, with most learning through self-study or social media rather than discipline-specific coursework.

Conclusion and Implications: Findings suggest that social science students are using AI in selective but uneven ways, while navigating it with limited formal preparation, substantial uncertainty about ethical and effective use, and considerable concerns about its implications for learning, misinformation, and humanistic dimensions of professional practice. The findings point to a need for discipline-specific AI education that moves beyond informal self-learning to provide clearer guidance, stronger ethical competency, and structured opportunities for dialogue about responsible use.