A new artificial intelligence simulation platform aims to transform how social workers are trained, offering students realistic, hands-on experience with lifelike virtual clients before they ever enter the field.
The platform, called Empathy Helper, was developed by a multidisciplinary team of social work educators, mental health professionals, and AI experts. Ryan Lindsay, a professor of practice at the Brown School at Washington University in St. Louis, and chair of the mental health concentration in the Master of Social Work program, co-led the project with Ruopeng An, a former Brown School faculty member and now the Constance and Martin Silver Endowed Professor in Data Science and Prevention at New York University’s Silver School of Social Work.
Empathy Helper uses advanced large language models and Retrieval-Augmented Generation (RAG), a technique that improves the accuracy and reliability of generative AI, to create dynamic, emotionally responsive simulated clients. Developers say the platform helps social work students build communication skills in a safe, consistent, and scalable environment.
“Our goal is to close the gap between classroom learning and social work practice,” Lindsay said. “Empathy Helper gives learners the chance to build confidence and competence through immersive practice, anytime, anywhere.”
The need for such training tools is growing. The U.S. Bureau of Labor Statistics estimates 751,000 social workers were employed nationwide as of 2023. Meanwhile, the Council on Social Work Education reported more than 140,000 students were enrolled in social work programs during the 2022–23 academic year, highlighting the need for effective, scalable training solutions.
“Empathy Helper is uniquely positioned to meet this need, offering a specialized platform tailored to the real-world demands and regulatory standards of social work education,” Lindsay said. He noted that Empathy Helper is the only purpose-built AI tool for the social work sector and will fill a significant gap in learning and assessment of social work competencies.
Launched in 2023, the project grew out of frustration with existing simulation tools in social work education. Initial seed funding was provided by the Brown School’s strategic initiative on Data Science and Technology to Advance the Social Good. More recently, additional support came from the WashU Gap Fund, operated by the Office of Technology Management (OTM). This new funding will support further development of the platform and establish pilot partnerships with leading social work schools.
Early results of the Empathy Helper platform have been encouraging. A 2024–2025 pilot study involving MSW students and clinical social work professionals showed notable improvements in both student self-efficacy and observable skills. The pilot study demonstrated high user satisfaction, ease of use and significant potential for scalability.
Participants shared the following feedback:
- “Using Empathy Helper was so much better than all the role-playing we had to do in our Practice 1 class. So much easier than trying to pretend like a client you know nothing about.”
- “I felt like I was talking to a real client.”
- “I really liked how I was able to take my time, use the resources the instructor gave me, and then respond. This helped me to learn it in a really in-depth way.”
Preliminary findings from the pilot study have been submitted for publication to the Journal of Social Work Education and are currently under review.
Lindsay and An will present on Empathy Helper at the 2025 Council on Social Work Education (CSWE) Annual Conference in Denver this October. Their workshop, “Empathy Helper: Simulated Learning for Inclusive Social Work Education,” will offer participants hands-on experience with the platform’s interactive client simulations and explore strategies for aligning simulated practices reflective of disability justice principles into social work curricula.