[Paper Review] Design Guidance Towards Addressing Over-Reliance on AI in Sensemaking

Yihang Zhao, Wenxin Zhang+3 more|arXiv (Cornell University)|Mar 9, 2026

Team Dynamics and Performance0 citations

TL;DR

The paper argues for integrating GenAI into group awareness tools (GATs) in a way that preserves implicit guidance to foster autonomous sensemaking, presenting a design space with three core considerations for presentation, deployment, and interaction.

ABSTRACT

Sensemaking in collaborative work and learning is increasingly supported by GenAI systems, however, emerging evidence suggests that poorly designed GenAI systems tend to provide explicit instruction that groups passively follow, fostering over-reliance and eroding autonomous sensemaking. Group awareness tools (GATs) address this challenge through implicit guidance: rather than instructing groups on what to do, GATs externalize observable collaboration data through visualizations that reveal differences between group members to create cognitive conflict, which triggers autonomous elaboration and discussion, thereby implicitly guiding autonomous sensemaking emergence. Drawing on an initial literature search of existing GAT systems, this paper explores the design of GenAI-augmented GATs to support autonomous sensemaking in collaborative work and learning, presenting preliminary design principles for discussion.

Motivation & Objective

• Motivate the problem of over-reliance on AI in sensemaking within collaborative work and learning contexts.
• Propose a design space for GenAI-augmented GATs that maintain implicit guidance rather than explicit instruction.
• Identify key design considerations for deploying, presenting, and interacting with GenAI-generated awareness information in GATs.

Proposed method

• Review and synthesis of existing GAT systems from ACM DL, IEEE Xplore, and Scopus, complemented by backward snowballing.
• Analysis of how GATs generate, present, and support exploration of group awareness information.
• Derivation of three consistent design considerations for integrating GenAI into GATs to preserve implicit guidance.

(a) Without GenAI integration (reproduced from (Zheng et al. , 2025a ) , translated into English with minor visual revisions).

Experimental results

Research questions

• RQ1 How can GenAI be integrated into GATs to provide implicit guidance that supports autonomous sensemaking?
• RQ2 What are the design considerations for deployment, presentation, and interaction when GenAI augments GATs?
• RQ3 How can GenAI enrich awareness information without turning into explicit instructional content?

Key findings

• GenAI should be used for qualitative interpretation of unstructured content rather than relying solely on rule-based quantitative signals.
• GenAI-generated awareness information should augment primary quantitative representations by surfacing semantic differences that foster cognitive conflict.
• Interaction techniques are needed to allow groups to explore GenAI insights and examine underlying evidence to trigger autonomous elaboration.
• A hybrid architecture combining rule-based processing for quantitative signals with GenAI for qualitative interpretation is preferable to end-to-end AI pipelines.
• GenAI-augmented visualizations should preserve implicit guidance and not replace traditional awareness cues.

(b) With GenAI integration.