Title:

LLM-Supported Decisions

This project investigates how AI assistance influences user attention and decision-making during online safety judgments. A browser-based experimental platform was developed using webcam eye-tracking to measure where users look while evaluating the safety of website URLs. Participants completed two counterbalanced task runs—one with AI-generated guidance and one without—while gaze data and behavioral responses were recorded in real time. The system captures attention allocation across defined interface regions and synchronizes gaze with user actions to enable detailed analysis. Results from a pilot study indicate that AI assistance alters visual attention patterns, increasing focus on guidance-related content, while also impacting decision speed and accuracy. These findings suggest that AI tools can meaningfully shape how users process information, with implications for interface design, cybersecurity education, and human-AI collaboration. Future work will expand the participant pool and refine AI feedback strategies to better understand how assistance can improve user outcomes without over-reliance.

Name

Comment*

Comments are viewable only by submitter

Captcha*