<p>The rapid spread of multimedia content on social media platforms such as TikTok presents significant challenges for detecting disinformation. Traditional detection approaches often focus on text-based analysis or multimodal fusion without fully addressing key issues such as model explainability, bias in detection mechanisms, and personality traits associated with disinformation spreaders. This study proposes a novel hybrid intelligence framework that integrates deep learning and fuzzy logic to enhance the detection of suspected disinformation in TikTok videos. The approach combines a multimodal feature analyser with a fuzzy logic-based detector to model behavioural traits across text, audio, and video. A key innovation lies in the use of a hierarchical fuzzy system that linguistically describes behavioural dimensions, enabling explainable and human-like reasoning in the detection process. Unlike existing studies, this framework incorporates explainability mechanisms to present transparent insights in the detection process and personality trait analysis to improve disinformation suspicion profiling. Two experiments were conducted: one evaluating disinformation within specific contexts and another assessing the scalability of the model across diverse topics. The results demonstrate that the proposed approach effectively identifies disinformation suspicion while generating high-quality, well-structured reports that ameliorate interpretability and trust in artificial intelligence based detection systems.</p>