Designing Inclusive Generative Artificial Intelligence-Based Conversational Agents with and for Intellectually Disabled Individuals

Generative AI-based conversational agents (CA) have become increasingly widespread due to their broad, domain-independent capabilities, helping many individuals complete tasks more efficiently. However, without inclusive designs, individuals with intellectual and developmental disabilities (IDD) are often unable to benefit equally from these systems, reinforcing digital marginalization. To address this, we apply the echeloned design science research approach and synthesize insights from a structured literature review, 19 interviews, and three focus groups with individuals with IDD. This results in five design requirements and 12 design principles for inclusive GenAI-based CA design. Thus, we contribute to inclusive IS design research by advancing the emerging body of knowledge on inclusive design for agentic systems, specifically for individuals with IDD. Practically, our findings provide guidance for CA developers and providers, supporting the creation of systems that foster independent and effective user engagement.