Adapting the algorithm: identity, agency, and the future of adaptation in the age of generative AI

Abstract This article examines the emergence of algorithmic adaptation in this pivotal transitional moment of generative Artificial Intelligence (AI), where texts, images, and sounds are recombined not through human intent but through probabilistic computation. Drawing on recent controversies in Vietnam—the replacement of actress Nguyễn Thúc Thùy Tiên in Chốt đơn (Deal!, 2025) and Vietnam Television’s AI-generated resurrection of musician Trịnh Công Sơn—it shows how AI adaptations unsettle core categories of adaptation studies: identity, agency, and authorship. While humanistic adaptations (including feminist and queer practices) affirm identity through intentional reinterpretation, AI adaptations tend to attenuate or reconfigure identity by recombining decontextualized data outside lived experiential frameworks. The article develops the Prompt Paradox, where human intent initiates but cannot determine outcomes, producing distributed agency across users, algorithms, and datasets. It also introduces algorithmic intermediality, in which hybridity arises at the computational level rather than through aesthetic negotiation, collapsing the line between original and variant. Framing these dynamics through posthuman theory (Hayles; Braidotti), queer adaptation studies (Demory), and debates on algorithmic colonialism (Birhane; Couldry and Mejias), the article builds on postmodern reconfigurations of adaptation studies that displace fidelity and originality. It argues that the algorithmic age demands a further theoretical shift—from human-centred notions of intertextuality and authorship to the recombinatory logic, post-originality, and distributed agency defining adaptation amid generative computation. Generative AI, the article concludes, does not end adaptation studies but repositions it critically at the intersection of technology, aesthetics, and cultural power in the posthuman era.