From Medical AI to Media Immunity: Adapting Explainable AI Diagnostic Frameworks for Disinformation Detection

Alsubai (2024) demonstrates the power of explainable AI (XAI) in medical diagnostics, using techniques like SHAP to make complex models interpretable for clinicians. Adapting this paradigm, this research proposes developing XAI-driven "media immune systems" that empower users to understand why news or social media content is flagged as disinformation. By leveraging local pattern analysis, multimodal data, and transparent feature attribution, the system would not only detect but explain risk factors directly to end-users—enhancing trust, literacy, and democratic engagement. This approach diverges from black-box detection systems and addresses the growing call (e.g., Sánchez Gonzales et al., 2024) for interpretable, actionable AI in the fight against disinformation. The cross-pollination from medical AI injects a new perspective, potentially transforming how citizens and professionals interact with information integrity tools.

References:

1. Artificial intelligence and disinformation literacy programmes used by European fact-checkers. Hada M. Sánchez Gonzales, María Sánchez González, Marián Alonso-González (2024). Catalan Journal of Communication & Cultural Studies.
2. Transfer learning based approach for lung and colon cancer detection using local binary pattern features and explainable artificial intelligence (AI) techniques. Shtwai Alsubai (2024). PeerJ Computer Science.