Logic, Dilemmas, and Pathways of Generative-AI-Driven Development of Publishing Cultural and Creative Products

In 2025, China's domestic book retail market reached a total print-run revenue of 110.4 billion yuan, representing a year-on-year drop of 2.24%. By contrast, publishing cultural and creative products achieved remarkable market expansion. The surging cultural derivative revenue of Shanghai Book Fair and viral literary peripherals such as Lu Xun-themed merchandise prove that publishing cultural creations have stepped out of scattered pilot attempts and entered a new phase of large-scale, branded operation. Even so, the sector is plagued by prominent bottlenecks including homogenized product forms, lengthy development cycles and monotonous marketing strategies. Fundamentally, traditional development modes relying on manual creation and linear copyright licensing cannot satisfy modern consumers' demands for personalization, experiential consumption and rapid iteration, failing to form a closed loop covering IP content excavation, product implementation and long-term monetization. Driven by the national "AI Plus" initiative set out in the country's development plans, generative artificial intelligence (GenAI) provides innovative technical approaches to tackle these industrial obstacles. This article elaborates on how GenAI acts as an innovation engine for publishing cultural creations across three layers. In content production, GenAI converts conventional linear workflows into collaborative brainstorming, lowering time and technical thresholds for multi-modal derivatives including audio books, dynamic posters and immersive interactive works. In marketing, big data profiling and digital human interaction transform extensive mass publicity into personalized targeted distribution, fostering lasting emotional connections between audiences and book IPs. In industrial collaboration, GenAI eliminates cross-media technical barriers to build integrated product matrices combining physical goods, digital assets and offline immersive spaces. Nevertheless, GenAI brings hidden risks termed "reconstruction myths". Shallow technical application weakens creators' subjectivity and dilutes cultural connotations; algorithmic recommendations create filter bubbles and cut off profound emotional resonance between readers and texts; insufficient industrial governance results in ambiguous copyright ownership and unequal resource allocation, with large tech giants and publishing groups dominating core industrial resources while small and medium publishers get marginalized. Centered on value creation, this research puts forward a three-pronged development framework. Generative value facilitates in-depth human-machine synergy by integrating algorithmic explicit knowledge and editors' tacit cultural experience. Relational value constructs embodied immersive reading scenarios to break algorithmic narrowness and restore deep textual communication. Governance value introduces traceable blockchain copyright registration, revenue-sharing industrial alliances and impartial public evaluation mechanisms to balance commercial gains and cultural public interests. Ultimately, GenAI is supposed to assist cultural inheritance and in-depth reading instead of merely boosting production efficiency, enabling publishing cultural creations to shoulder the responsibility of passing down civilizations amid digital transformation.