Paper Number
ECIS2026-1658
Paper Type
CRP
Abstract
Information systems (IS) research has largely treated artificial intelligence (AI) agents as singular entities, which has obscured key challenges these agents encounter when they interact with other systems to complete their tasks. To remedy this issue, we offer a framework to study AI agents as autonomous systems of multiple IS. This view reveals conceptual detail underlying AI agents and points to important gaps that AI agents encounter and must overcome when they interact with other systems. Building on representation theory, we discuss three categories of gaps associated with the deep structure, the surface structure, and the physical structure of the systems involved in an AI agent’s task performance. Overcoming these gaps presents new challenges to AI agents themselves, their design, and their use in organizational practice. Our framework offers a foundation to address these challenges, and we use our framework to derive avenues for future research on AI agents.
Recommended Citation
Gerlach, Jin P. and Gnewuch, Ulrich, "Autonomous Multi-IS Systems: A Framework For Studying Artificial Intelligence Agents" (2026). ECIS 2026 Proceedings. 7.
https://aisel.aisnet.org/ecis2026/is_adopt/is_adopt/7
Autonomous Multi-IS Systems: A Framework For Studying Artificial Intelligence Agents
Information systems (IS) research has largely treated artificial intelligence (AI) agents as singular entities, which has obscured key challenges these agents encounter when they interact with other systems to complete their tasks. To remedy this issue, we offer a framework to study AI agents as autonomous systems of multiple IS. This view reveals conceptual detail underlying AI agents and points to important gaps that AI agents encounter and must overcome when they interact with other systems. Building on representation theory, we discuss three categories of gaps associated with the deep structure, the surface structure, and the physical structure of the systems involved in an AI agent’s task performance. Overcoming these gaps presents new challenges to AI agents themselves, their design, and their use in organizational practice. Our framework offers a foundation to address these challenges, and we use our framework to derive avenues for future research on AI agents.
When commenting on articles, please be friendly, welcoming, respectful and abide by the AIS eLibrary Discussion Thread Code of Conduct posted here.