Abstract
Agentic artificial intelligence is transforming smart homes from reactive automation systems into increasingly autonomous environments capable of reasoning, coordination, and adaptive decision-making. While this shift promises greater efficiency and personalization, it also raises critical challenges related to accountability, transparency, and ethical governance. Existing smart home designs largely conceptualize autonomy as a technical capability rather than a governed socio-technical phenomenon. Drawing on Service-Dominant Logic and the Hybrid Intelligent Service Ecosystem framework, this study proposes the Agentic AI Home System (AAIHS) as a design artifact for responsible agentic smart homes. The architecture integrates three design mechanisms, configured agency, computable institutions, and reflexive learning, to enable appropriate autonomy, institutional alignment, and accountable adaptation in domestic AI environments. Adopting a Design Science Research approach, the study presents a layered system architecture and human–AI interaction design that balance machine initiative with human oversight while ensuring transparency and institutional governance. The study contributes to research on agentic information systems by operationalizing ecosystem-level governance mechanisms for autonomous systems and provides design guidance for developing responsible, transparent, and collaborative smart home technologies.
Keywords:
- Keyword: Agentic AI
- Keyword: Smart homes
- Keyword: Hybrid Intelligent Service Ecosystems
- Keyword: Responsible AI Design
- Keyword: Configured Agency
- Keyword: Human–AI Collaboration
How to Cite:
Bokka, A., Liang, S., Huang, Z., Ning, X. & Yu, Y. A., (2026) “Designing Responsible Agentic AI for Smart Homes: A Hybrid Intelligent Service Ecosystem Perspective”, Journal of Intelligent and Sustainable Systems (JISS) 2(1).