The increasing amount of digital information in smart cities introduces challenges related to misinformation, which can compromise decision-making processes and diminish citizen trust. This research introduces an innovative framework for the real-time detection of disinformation, utilizing advanced artificial intelligence (AI) techniques, data interoperability, and Linked Data. The system utilizes Natural Language Processing (NLP) models, such as BERT and Support Vector Machines (SVM), to classify and assess data from physical sensors, social media, and diverse urban sources. Moreover, the system employs semantic ontologies and interlinking techniques to enhance and verify data in real-time, thus increasing the transparency and reliability of public information. The framework’s major objective is to improve urban information quality, reduce misinformation spread, and enable smooth and scalable integration of varied data. The expected outcomes include better accuracy in predicting false information, automated and open data governance, and the chance to improve the system using new technologies like blockchain for decentralized validation and federated learning to keep making predictive models better. The simulations and testing conducted on real data will demonstrate the framework’s effectiveness within smart cities, positioning it as a crucial tool for enhancing informational transparency in urban environments.