Paper Number
ECIS2026-2912
Paper Type
CRP
Abstract
Existing descriptions of AI roles in human‑AI collaboration remain inconsistent and often contradictory. For example, some define AI assistants as strictly reactive, while others attribute them with proactive behavior. To address this lack of clarity, we develop a taxonomy that systematically characterizes human‑centered AI systems based on their representation and capabilities. The taxonomy was created through an iterative process combining conceptual reasoning and empirical analysis of literature artifacts and comprises 10 dimensions with 37 characteristics. The findings show that most systems maintain human decision authority while varying in autonomy, personification, and transparency. Organizations increasingly employ anthropomorphic designs and natural language communication to strengthen social presence and improve task performance. The taxonomy provides a structured framework for defining AI roles within collaborative settings and offers orientation for researchers seeking conceptual clarity as well as guidance for practitioners designing effective human‑centered AI systems.
Recommended Citation
Nebel, Maximilian; Speckmann, Pauline; Ciftci, Seyyid Ahmed; and Janiesch, Christian, "Standing On The Shoulders Of Paperclips: A Taxonomy Of Human-Centered AI Systems" (2026). ECIS 2026 Proceedings. 24.
https://aisel.aisnet.org/ecis2026/litrev/litrev/24
Standing On The Shoulders Of Paperclips: A Taxonomy Of Human-Centered AI Systems
Existing descriptions of AI roles in human‑AI collaboration remain inconsistent and often contradictory. For example, some define AI assistants as strictly reactive, while others attribute them with proactive behavior. To address this lack of clarity, we develop a taxonomy that systematically characterizes human‑centered AI systems based on their representation and capabilities. The taxonomy was created through an iterative process combining conceptual reasoning and empirical analysis of literature artifacts and comprises 10 dimensions with 37 characteristics. The findings show that most systems maintain human decision authority while varying in autonomy, personification, and transparency. Organizations increasingly employ anthropomorphic designs and natural language communication to strengthen social presence and improve task performance. The taxonomy provides a structured framework for defining AI roles within collaborative settings and offers orientation for researchers seeking conceptual clarity as well as guidance for practitioners designing effective human‑centered AI systems.
When commenting on articles, please be friendly, welcoming, respectful and abide by the AIS eLibrary Discussion Thread Code of Conduct posted here.