AI is driving a shift in the media content industry from being viewed as a consumer good to a dual-pricing model based on data and IP assets, with both traditional publishing and online literature IP set to benefit from this revaluation.
On one hand, traditional publishing copyrights, due to the scarcity of high-quality material, their verifiable ownership, and capacity for sustainable updates, are transforming from a cost center into a vital production asset in the AI era.
On the other hand, online literature IP has gained increased value as AI lowers the production barriers for short dramas, comic adaptations, and animations, highlighting the scarcity of premium story cores, world-building, and character assets.
Core Revaluation Rationale for Traditional Copyrights
The primary reassessment logic for traditional copyrights centers on the escalating "corpus value" of high-quality content data.
First, the market for trading high-quality training data overseas has matured. Projections indicate the global related licensing market could reach $22.6 billion by 2034, with landmark commercial deals like Anthropic's $1.5 billion class-action settlement, Taylor & Francis's $10 million agreement, and HarperCollins securing $5,000 per old book license fee, predominantly based on proprietary licensing, validating the strong bargaining power derived from copyright scarcity and specialization.
Second, while China's market currently focuses on localized private data deployment and self-developed models, policy is actively promoting the construction of high-quality datasets. Initiatives such as the "Mainstream Value Corpus Ecosystem Alliance" led by People's Daily Online and the National Data Administration's explicit proposal to build a "Token-based dataset value system" provide direction for the assetization and rights confirmation of data.
Finally, regarding implementation pathways, there remains room for "data resource capitalization on balance sheets" within the media sector. The current situation where the A-share media sector reported only 66 million yuan in "intangible assets - data resources" for 2025 is expected to change. A vast amount of previously uncapitalized internal copyrights and high-quality corpora are poised for expansion, leading to a reshaping of the valuation framework for publishing enterprises.
Value Concentration Towards IP Holders and Distribution Platforms
Against the backdrop of a production cost revolution and supply explosion driven by AI multimodal technology, there is a stronger outlook for value concentration towards IP owners and distribution platforms, with commercial space opening for IP holders and platforms requiring new AI-driven recommendation and matching engines.
First, the domestic IP market exhibits a trend of supply-demand resonance and diversified commercialization efforts. The scale of online literature IP adaptations reached 367.61 billion yuan in 2025. With micro-drama users exceeding 850 million and the overall short drama market surging to 108 billion yuan, over 80% of funded comic-drama projects rely on clear authorization from top online literature IP.
Second, the supply explosion imposes higher demands on platform "content-user" matching, with upstream IP content owners and downstream distribution platforms poised to capture the greatest value. AI has significantly lowered the threshold for content industrialization, compressing the production cost of a standard AI-simulated human short drama to 10,000-50,000 yuan per title—a reduction of 80%-90%—and shortening the production cycle to 1-2 weeks. Reviewing historical major productivity shifts in the content industry reveals that upstream IP content and distribution platforms consistently accumulate value, a trend expected to intensify in the AI era of supply proliferation.
Recommended Investment Approaches
It is recommended to focus on two core investment themes: "High-Dividend Publishing + Data Asset Revaluation" and "Online Literature IP + AI Content Industrialization," to comprehensively capture the dual valuation elasticity from stable dividends and earnings growth.
On one hand, traditional publishing enterprises hold significant advantages such as high barriers via education licenses, substantial market share in textbook distribution, and deep cultural corpora, while also providing a solid safety cushion through ample cash flow dividends. Companies to watch include Wanxin Media, Phoenix Publishing & Media, Zhongnan Media, Zhongyuan Media, and Xinhua Winshare.
On the other hand, online literature IP platform companies possess the most aggressive core capabilities in the digital entertainment era through a complete ecosystem of "IP reserves — industrialized adaptation — platform distribution — commercial realization." Companies to watch include Chinese All Digital Publishing, which has accumulated over 600TB of high-quality multimodal data resources; iReader Technology, which has elevated its short drama business to its largest segment; and China Literature Ltd., which produces over 800,000 works annually. These companies are leveraging massive resource libraries and AIGC toolchains to pioneer the realization of broad commercial potential in overseas expansion and multi-format development.
Risk warnings include potential underperformance in copyright licensing, data resource balance sheet capitalization, and related policy implementation; slower-than-expected iteration and commercialization of AI multimodal models; and increased content regulation, changes in platform rules, or intensified copyright disputes.
Comments