This study is motivated by fundamental changes in documentary cinema resulting from the emergence of artificial intelligence (AI) technology. Documentary films, which have traditionally been understood as a medium for recording reality, are now moving toward a new form as a medium for algorithm-based visual reconstruction. The main problem of this study lies in how visual language and algorithmic narrative in the AI documentary film Nusantara construct historical reality, cultural memory, and Indonesian identity amid the development of digital cinema. This study aims to analyze the visual language, algorithmic narrative, and forms of mediamorphosis in the AI documentary film Nusantara, which received the Best AI Documentary award at Cannes 2025. This research employs a descriptive-interpretive qualitative method with a case study approach. The research data consist of scenes, visual composition, color, lighting, image movement, narrative text, cultural symbols, historical setting, editing rhythm, and forms of AI-based visual reconstruction. Data were collected through audiovisual observation, documentation, note-taking, transcription, and scene segmentation, while the data were analyzed using visual semiotic analysis, narrative analysis, and mediamorphosis analysis. The findings present Nusantara as a representation of the transformation of documentary cinema from the recording of reality to algorithmic reconstruction. The film’s visual language is constructed through the figure of Gajah Mada, symbols of Majapahit, the Palapa Oath, dramatic colors, contrasting lighting, and colossal scenes. Its algorithmic narrative appears through the use of AI in constructing historical memory, expanding visual space, and presenting Indonesian cultural identity to a global audience. This study strengthens the position of AI documentary film as a new creative space for Indonesian cinema, while also demanding historical accuracy, technological transparency, and ethical responsibility in cultural representation.