Guosen: AI game supply accelerates, industry chain value extends toward generation, distribution, and premium content.

Guosen released a research report stating that industry opportunities are expected to unfold along the full chain of "models and toolscontent supplyplatform distributionR&D operations/IP," and investors should watch opportunities across the entire industry chain. 1) Upstream models and generative tools will benefit first from expanding creative demand; midstream traffic platforms will absorb newly added content and improve distribution and monetization efficiency; downstream game development companies will use AI to boost production efficiency and rely on productization, long-term operations, and IP capabilities to complete the final value conversion. China has already formed a relatively complete industrial foundation spanning AI creation tools, mini-game platforms, and game R&D and IP operations. 2) The bank remains bullish on large-model vendors with leading capabilities and continuously improving inference efficiency and multimodal capabilities; tool chains and platforms that are the first to achieve AI game generation implementation and build productization capabilities; platform-type companies with scarce traffic entry points, recommendation distribution, and mature IAA/IAP systems; and leading game CPs with top-tier IP, industrialized R&D, and long-term operations capabilities. Guosen's main points are as follows: Meta launches Horizon Create/Studio, with AI games further moving from assisted development toward full content generation Meta Horizon Create/Studio supports users in generating complete 2D/3D games through natural language, including elements such as gameplay mechanics, progression systems, art, and multiplayer connectivity, and directly connects to the Facebook, Instagram, and Horizon distribution systems. The bank believes this product launch marks Meta's completion of the platform of "AI generationcontent distributionuser reach," and that AI's application in the gaming industry is beginning to extend from a single-point production tool to a complete content production and distribution system. Gaming tool AI significantly lowers the threshold for game development, and game content supply is expected to enter a phase of accelerated growth 1) Roblox Cube generates more than 60,000 3D assets per day, with nearly 1,400 games using Cube-generated content daily; in 2026H1, TapTap added 5,002 playable games, up 844% year over year, of which more than 4,000 were made with the help of AI. 2) AI is significantly improving game development efficiency, expanding the range of creators, and driving growth in game content supply. As capabilities such as multimodality, 3D, and code generation further improve, AI is expected to gradually deepen from assisting with assets and partial segments into gameplay, levels, and complete game generation gaming tools. After content supply grows through gaming tools, the competitive focus of the gaming industry is expected to further extend toward distribution efficiency and commercialization capabilities 1) As the threshold for game production declines and content supply is no longer the only constraint, the importance of completing user reach, content discovery, retention, and commercialization continues to rise. 2) Meta this time directly connects game generation to social feeds, while WeChat mini-games and Douyin mini-games in China have already formed relatively mature "play instantly + IAA/IAP" business models. The combination of new AI supply and mature platform traffic systems is expected to further improve game content distribution efficiency and open up user time and commercialization space. Risk warnings: model and AI game generation capability improvements fall short of expectations; AI-generated game content quality, user retention, and commercialization performance fall below expectations; game content supply grows faster than user demand, leading to intensified competition; platform distribution and commercialization policies change; game companies' AI R&D investment and cost reduction and efficiency improvement results fall short of expectations; data security, content copyright, minor protection, and other regulatory policy changes, etc.