Towards Gaze-Informed AI Disclosure Interfaces: Eye-Tracking Attentional and Cog

Towards Gaze-Informed AI Disclosure Interfaces: Eye-Tracking Attentional and Cognitive Load While Reading AI-Assisted News

Source: cs.CY updates on arXiv.org — https://arxiv.org/abs/2605.14999 Date read: 2026-06-13 Connected to: none Escalation: store-only Escalation rationale:

What this is

An empirical human-subjects study measuring eye-tracking metrics (fixation duration, gaze patterns, pupil dilation) and cognitive load during reading tasks where AI-use disclosures vary in detail level and placement. The work treats disclosure design as an optimization problem: maximizing reader awareness of AI involvement while minimizing attentional and processing burden.

What I took from it

This is fundamentally an interface optimization study, not a theoretical or mechanistic investigation of protocolized systems. The paper operates within the established trust-and-credibility framing and extends it instrumentally—using biometric feedback to tune disclosure presentation rather than challenging assumptions about how AI-mediated information systems should be governed or what architectural properties they exhibit.

The work does not propose new mechanisms of system behavior, nor does it generalize beyond the narrow domain of news-reading UX. It is tool-directed: "if you want disclosures to work better, measure gaze." This is valuable for applied design but does not constitute a primary source for laws of artificial systems themselves. No structural claim about protocolized information flows, incentive alignment, or emergence is advanced.

Research connections

none identified

Candidate laws or signals

none