Session: A Stakeholder-Driven Approach for Ethical AI Integration into Social Work Practice and Education at a Public University In Southwest, USA. (Society for Social Work and Research 31st Annual Conference)

AI large language models (AI LLM) have dramatically changed the ways we work in social work as well as higher education. There is limited evidence illuminating how social workers should engage in conversation about AI LLM’s application in practice. This symposium aims to share findings from a mixed-methods study, including focus group interviews and surveys, conducted with key stakeholders in a social work program at a large public university in the Southwestern United States that offers both online and in-person BSW and MSW programs. A school-level committee was established to promote evidence-based dialogue on the role of AI LLM in social work practice and education. Drawing on data collected during the 2025–2026 academic year, this symposium presents perspectives from students, faculty, staff, and practicum supervisors.

The first study used focus group data to document use cases of AI LLM in social work. Findings highlight three primary functions: documentation support, communication support, and expanded access to information. Despite benefits, participants raised concerns about confidentiality, accuracy, and the compatibility of AI with professional values and ethics.While AI LLM are intuitive and accessible, its adoption is shaped less by technological barriers and more by ethical concerns and value alignment.

The second study examines how AI LLM use aligns with the six core values of the National Association of Social Workers (NASW). Participants identified both opportunities and tensions across all six core values. While AI LLM can enhance efficiency, concerns raised as to potential loss of human connection, threats to confidentiality, and the need for transparency and informed consent. Findings suggest the six core values are promising as a guiding framework for AI LLM adoption.

The third study developed and validated the AI Social Work Task Application Scale (AISWTAS), a 24-item measure assessing willingness to use AI LLM across task-specific domains. Results demonstrate strong reliability and validity, with higher scores among frequent AI LLM users. The scale provides a practical tool for assessing task-level adoption and offers a structured framework for finding social work jobs subject to AI LLM’s application.

The fourth study applies the Technology Acceptance Model 2 (TAM2) to examine predictors of AI adoption, incorporating ethical concerns as a central factor. Adoption is driven primarily by personal values, particularly “AI guilt”, which reflects discomfort with delegating professional tasks to AI as well as an image associated with using AI LLM within an organization. Open dialogue and structured guidance on ethical AI use appear important before leadership implements technical support within an organization.

Findings from four presentations offer a big picture that while AI LLM are accessible and potentially making practice efficient, their adoption in social work is shaped primarily by ethical considerations and value alignment rather than technical barriers. This symposium highlights the need for structured training, clear guidelines, and ongoing dialogue to support ethical, responsible, and value-coherent integration of AI LLM in social work practice and education.