The preservation of intangible cultural heritage (ICH) has gradually shifted to digital innovations in the era of expanding information technologies (IT). This study uses brief video communication to investigate the function of artificial intelligence (AI) in ICH protection. A social networking platform is connected with the AI-driven method. This study comprises instances of short videos and concentrates on essential components like musical instruments and traditional music. Next, the linguistic (text) and visual aspects of the short videos are evaluated using computer vision techniques like You only look once (YOLO) network and natural language processing (NLP) approaches. We use AI-driven analysis to find trends in the data, and we also use qualitative analysis to present it. The random forest (RF) model is developed to assess the heritage transmission level in the videos based on factors like security and cultural importance. The outcomes of the RF model are analyzed in terms of precision, accuracy, and f-measure metrics. Additionally, performance comparisons of the allocation of traditional musical instruments under different production types are carried out, and outcomes are also obtained visually using data visualization tools. This study finally shows how AI can be used to enhance ICH safety through short video communication technology.