Context, Credibility, and Control: User Reflections on AI Assisted Misinformation Tools

TL;DR

This paper presents an interactive AI tool designed to improve user ability to identify misinformation on social media by providing contextual explanations, source aggregation, and debate-style interactions, enhancing critical thinking and trust.

Contribution

It introduces a novel collaborative AI interface with real-time explanations and debate features, demonstrating improved user engagement and perceived usefulness in misinformation detection.

Findings

79% of participants preferred debate mode over standard chatbots

Average usefulness rating of 4.6 out of 5 for multiple-source view

Dialogic AI systems can enhance media literacy and user trust

Abstract

This paper investigates how collaborative AI systems can enhance user agency in identifying and evaluating misinformation on social media platforms. Traditional methods, such as personal judgment or basic fact-checking, often fall short when faced with emotionally charged or context-deficient content. To address this, we designed and evaluated an interactive interface that integrates collaborative AI features, including real-time explanations, source aggregation, and debate-style interaction. These elements aim to support critical thinking by providing contextual cues and argumentative reasoning in a transparent, user-centered format. In a user study with 14 participants, 79% found the debate mode more effective than standard chatbot interfaces, and the multiple-source view received an average usefulness rating of 4.6 out of 5. Our findings highlight the potential of context-rich, dialogic AI systems to improve media literacy and foster trust in digital information environments. We argue that future tools for misinformation mitigation should prioritize ethical design, explainability, and interactive engagement to empower users in a post-truth era.