Audience Behavior Transformation in the Era of Streaming and Artificial Intelligence: Perspectives from Film, Media, and Communication

Abstract

Streaming platforms and artificial intelligence (AI) have jointly reorganized how audiences discover, watch, and discuss film and media content, yet most scholarship still treats recommendation algorithms, viewing habits, and platform economics as separate lines of inquiry. This article synthesizes twenty-five peer-reviewed and conference sources published mainly between 2021 and 2026 to examine how AI-driven personalization, generative tools, and streaming infrastructures are reshaping audience behavior across film, broadcasting, and social media contexts. Using a narrative literature synthesis of English- and Indonesian-language reference corpora supplemented by additional database searches, the study traces three converging shifts: from scheduled, appointment-based viewing to on-demand and algorithmically curated consumption; from anonymous mass audiences to data-legible, hyper-personalized users; and from passive spectatorship to participatory, algorithm-aware engagement through comments, ratings, and generative co-creation. The review proposes an original AI-Mediated Audience Transformation (AI-MAT) framework, positioning audiences along a continuum from default viewers to algorithmically negotiated participants. Findings indicate that while personalization increases engagement, satisfaction, and retention, it simultaneously raises concerns about algorithmic bias, filter bubbles, transparency, and diminished audience autonomy. The article concludes that film, media, and communication scholarship must integrate audience-reception theory with computational media studies to adequately theorize AI-mediated viewing publics, and outlines directions for future empirical and cross-cultural research.