Session: AI, Policy, and Equity In Suicide Prevention: Advancing Rigorous Social Work Research for the Common Good (Society for Social Work and Research 31st Annual Conference)

220 AI, Policy, and Equity In Suicide Prevention: Advancing Rigorous Social Work Research for the Common Good

Schedule:
Saturday, January 23, 2027: 9:45 AM-11:15 AM
Golden Gate 7 (Hilton San Francisco Union Square)
Cluster: Health
Symposium Organizer:
Amanda Ritchie, New York University
Discussant:
Ruopeng An, PhD, New York University
Suicide prevention is an urgent public health and social justice issue, and early detection remains a critical gap in research and practice. In the context of a rapidly evolving social and technological landscape, advancing equitable and effective approaches to suicide prevention is essential to promoting collective well-being and strengthening the common good. Emerging methods leveraging artificial intelligence (AI), big data, and other computational approaches are increasingly shaping social work research and offer new opportunities to identify risk earlier and intervene more effectively. However, these innovations must be developed and applied in ways that are equitable, inclusive, and responsive to the needs of marginalized populations. This symposium showcases three student-led projects from the NYU Silver School of Social Work and its C+M Center, each focused on early detection of suicide risk, methodological innovation, and reducing disparities.

The first paper uses computational policy analysis and causal inference methods to examine the effects of anti-LGBTQ+ legislation on suicide risk among youth. By identifying how specific domains of legislation differentially impact mental health outcomes, this study highlights the role of structural determinants in shaping suicide risk and offers actionable insights for policymakers and advocates. The findings underscore how rigorous research can inform more equitable, accountable, and democratic policy-making.

The second paper presents a multi-agent large language model (LLM) system that integrates social media data with a clinically grounded taxonomy to detect early signals of suicide risk. The system improves both detection accuracy and transparency in decision-making, offering a scalable, real-world tool for social workers and other professionals. By enabling earlier identification of individuals at risk, this approach has the potential to expand access to care and strengthen prevention efforts across diverse settings.

The third paper proposes a culturally grounded framework for applying AI models to detect suicide risk among Asian American populations. The authors identify how existing natural language processing (NLP) models privilege Western expressions of distress and may fail to capture culturally nuanced help-seeking behaviors. By advancing a more inclusive and culturally responsive framework, this work promotes more equitable detection and intervention and ensures that marginalized voices are more accurately represented in digital mental health systems.

Taken together, these papers advance early detection of suicide across multiple levels - policy, technology, and culture - through the use of innovative and rigorous methodologies. They demonstrate the capacity of social work research to bridge systems, inform policy, and reduce disparities in access to care. This symposium highlights how rigorous, inclusive inquiry can advance social work research in service of democracy by promoting equitable participation in mental health systems and strengthening the common good through prevention and well-being. The session will also invite discussion on the ethical and practical implications of technology-driven approaches to suicide prevention and their role in shaping a more just and inclusive future.

* noted as presenting author
Which Anti-LGBTQ+ Bills Hurt Most? a Computational Policy Analysis of Legislative Domains and Adolescent Suicidal Ideation
Dget Downey, MSW, New York University; Madison Kitchen, EdM, MSW, Boston University; Jacqueline Cosse, MSW, New York University; Avery Hays, New York University; Jessica Fish, PhD, University of Illinois at Urbana-Champaign
Detecting Early Suicide Risk from Social Media with a Multi-Agent LLM System
Jingyi Huang, MA, New York University; Michael Lindsey, PhD, MSW, MPH, New York University; Ruopeng An, PhD, New York University
Understanding Asian American Help-Seeking through NLP: Towards Culturally Responsive Suicide Prevention
Xiaoran (Livia) Liu, MSW, New York University; Ruopeng An, PhD, New York University
See more of: Symposia