China Securities Co.,Ltd. October AI Monthly Report: Model Iteration and Investment Remain Strong, Agent Commercialization Continues to Materialize

China Securities Co.,Ltd. released a research report stating that overseas security discussions have not changed the high-frequency model iteration pace, and RSI is gradually becoming a consensus among leading model companies such as OpenAI and Anthropic; domestically, Alibaba (09988) and Z.AI completed financing of approximately HK$80 billion and US$5 billion respectively, with large-scale investment supporting model capability catch-up, while Tencent (00700) also increased overseas computing power procurement. On the commercialization front, overseas enterprise cooperation is deepening, ChatGPT advertising annualized revenue run rate reached US$1 billion, and Muse is broadening consumer monetization paths; domestic Agents are still expected to drive Token calls and a new round of ARR growth by increasing per-user task density. The overall trend of the domestic AI sector is upward, and it is recommended to focus on computing power services, domestic chips and super nodes, as well as B-end AI application vendors with scenario, data, and enterprise delivery capabilities. The main views of China Securities Co.,Ltd. are as follows: Overseas: Model capabilities continue to improve, enterprise deployment and consumer monetization are landing Security discussions have not changed the high-frequency model release pace. OpenAI and Anthropic successively updated GPT-6 Sol/Luna, Opus 5.5, and GPT-6.1 Sol, with the latter approaching the capabilities of GPT-6 Astra while costing only 1/5 as much. R&D automation has begun to feed back into model iteration. As of August, the proportion of R&D work led by Claude under human supervision rose to 26%. On the commercialization front, Anthropic deepened cooperation with Salesforce and Bain to push Agents into enterprise processes; ChatGPT advertising annualized revenue run rate reached US$1 billion, and Muse expanded personal task entry points, beginning to release consumer commercial value. Domestic: Financing supports model expansion, Agent tasks accelerate ARR upward Alibaba completed a placement of approximately HK$80 billion, Z.AI completed approximately US$5 billion in equity and debt financing, and Tencent also increased overseas computing power procurement. On the commercialization front, MiniMax's Token consumption in July reached 20 times that of January, and Z.AI raised its year-end ARR target from US$2.4 billion to US$3 billion. Competition in office Agents and the Harness ecosystem are still evolving. The next stage of growth depends more on effective task volume, usage retention, and paid conversion. Model capability improvements are expected to drive a new round of revenue expansion. Market review: Hardware and overseas software led the recovery Using the July 31 close as the base period, as of September 30, IGV and the Philadelphia Semiconductor Index rose 12.6% and 11.6% respectively; A-share communications and electronics rose 4.7% and 3.2%, while computers and media fell 5.2% and 3.2%. In terms of valuation, IGV's comprehensive PS-TTM was 8.59x, while SW computers and media were 2.80x and 2.68x respectively, down 10.1%, 21.6%, and 18.6% from the end of 2025. The growth in AI payments and data calls for overseas software provides support for valuation recovery, while the performance of the domestic application side remains relatively divergent. Subsequent pricing will place greater emphasis on customer procurement, sustained usage, and profit contribution. Enterprises with clear AI revenue and delivery capabilities are more likely to obtain valuation support. Risk warnings (1) AI industry commercialization landing falls short of expectations: At present, the commercialization models of AI products at various stages are still in the exploration stage. If the pace of advancement of products at various stages falls short of expectations, it may adversely affect the performance of related enterprises; (2) Market competition risk: Overseas AI vendors, relying on first-mover advantages and strong technological accumulation, occupy an advantageous position in competition. If domestic AI vendors' technology iteration falls short of expectations, their operating conditions may be affected; at the same time, many domestic enterprises have already invested in AI product R&D, and there may subsequently be risks of homogeneous competition, which may in turn affect the revenue of related enterprises; (3) Policy risk: The development of AI technology is directly affected by the policies and regulation of various countries. As AI penetrates various fields, the government may further introduce corresponding regulatory policies to regulate its development. If enterprises fail to adapt to and comply with relevant policies in a timely manner, they may face corresponding penalties or even be forced to adjust business strategies. In addition, policy uncertainty may also lead to errors in corporate strategic planning and investment decisions, increasing operational uncertainty; (4) Geopolitical risk: Amid fluctuations in the global geopolitical environment, especially US export restrictions on China may directly affect domestic enterprises' access to computing power chips, thereby affecting their product R&D and market competitiveness. At the same time, geopolitical risk may also cause AI products to face obstacles in expanding into overseas markets, affecting the revenue of related enterprises.