Core Insights - The competition in the AI agent sector is intensifying, with a focus on "machine execution capability" as AI agents evolve from passive responders to proactive collaborators in various fields [1] - Financial AI agents face unique challenges due to the industry's high-risk and heavily regulated nature, necessitating high reliability and accuracy [1] Financial AI Agents: Unique Characteristics and High Barriers - Financial AI agents must achieve 100% execution accuracy and seamless integration with complex systems to handle sensitive tasks like risk control and financial reconciliation [1] - The market for financial AI agents can be categorized into different "capability quadrants," highlighting the varying value propositions and applicable scenarios [1] Professional AI Agents: Breaking Through Complex Business Scenarios - Jinzhihui's AI Agent platform stands out by focusing on high-value, business-relevant deep empowerment, achieving significant automation in core business processes [2] - The platform enhances the "Reflection" capability, allowing for repeated validation of task results, ensuring precision in critical financial scenarios [2] - Jinzhihui's unique "Browser Use" model improves script development efficiency by over 70%, facilitating a shift from "how to execute" to "defining business" [2] Execution-Focused AI Agents: Meeting Specific Automation Needs - Douzi, developed by ByteDance, offers a low-code, visual workflow design, enabling quick assembly of marketing automation and intelligent customer service applications [4] - While ideal for marketing teams seeking rapid deployment, Douzi lacks the necessary accuracy and compliance for complex core business processes in regulated industries [4] Tencent Yuanqi: Leveraging WeChat Ecosystem - Tencent Yuanqi maximizes the WeChat ecosystem's potential, integrating various Tencent services for quick deployment of intelligent services in specific scenarios [6] - However, it does not meet the financial-grade requirements for handling complex core business processes [6] Zhiyu Qingyan: Expertise in Professional Text Understanding - Zhiyu Qingyan excels in deep semantic understanding and generation of professional texts, making it a valuable tool for research and legal institutions [7] - It faces limitations in linking complex systems and meeting the reliability and accuracy standards required in the financial sector [7] Baidu Lingjing: Intelligent Management of Enterprise Knowledge - Baidu Lingjing focuses on the intelligent management of enterprise knowledge assets, providing precise answers through smart retrieval [7] - It is crucial for large organizations aiming to integrate and activate scattered knowledge but is not suited for core business automation in finance [7] Dify: Open-Source LLM Application Development Framework - Dify offers a comprehensive toolchain for developing applications, appealing to teams with strong technical capabilities seeking customization [7] - However, it lacks the reliability and precision needed for financial-grade core business automation [7] AI Agent Selection: Linked to Enterprise Digital Maturity - The choice of AI agents is a significant decision tied to the level of digital maturity within an enterprise, impacting productivity liberation [9] - For enterprises with lower digital maturity seeking quick trials, execution-focused products may suffice, while those in regulated industries with higher maturity require robust solutions like Jinzhihui to enhance core competitiveness [9]

Core Insights - The article discusses the significant evolution of AI products in 2025, highlighting a shift from merely "talking" to "doing" [3][4] - The focus is on the transformation of interaction paradigms and the integration of AI into both digital and physical realms [5][6] - The article introduces the "AI 100" product list, categorizing AI products into flagship and innovative segments, along with five major application categories [6][9] Group 1: AI Product Development - AI products have shown differentiated growth across various sectors, with strong demand in general scenarios and AI efficiency, while AI life products are exploring better user experiences [14] - The common goal across all sectors is moving towards end-to-end delivery of productivity, shifting the value measurement from "how well it answers" to "how completely it delivers" [14][15] Group 2: Flagship AI Products - The "Flagship AI 100" and "Innovative AI 100" categories represent the strongest and most promising AI products, respectively [7][13] - The article outlines ten core tracks for AI applications, including AI smart assistants, AI agents, AI browsers, AI workstations, Vibe Coding, AI education, AI entertainment, AI health, multimodal creation, and AI consumer hardware [9][10] Group 3: AI Smart Assistants - AI smart assistants are the most traffic-intensive and revenue-near segment, evolving from answering questions to solving problems [16] - Top products in this category include: - Doubao from ByteDance, with over 57 million daily active users [18] - DeepSeek, known for its innovative interaction method that showcases AI reasoning [20] - Tencent Yuanbao, integrating various social networks for enhanced user experience [22] Group 4: AI Agents - AI agents have transitioned from mere conversational tools to executing tasks [23] - Notable products include: - Nano AI from 360 Group, which integrates over 80 large models for task execution [24] - Kouzi, a one-stop AI office space from ByteDance, automating complex workflows [26] - Xingliu, a new generation AI creation tool from Singularity Star, facilitating end-to-end creative processes [30] Group 5: AI Browsers - AI browsers are evolving from passive information displays to active task executors [32] - Key products include: - QQ Browser from Tencent, which integrates AI capabilities to understand user intent [33] - Quark from Alibaba, combining search, reading, and creation functionalities [36] - Fellou, focusing on a unified search and task experience [40] Group 6: AI Workstations - The competition in AI workstations has shifted from the number of features to complete workflow integration [41] - Leading products include: - Baidu Wenku, transforming from a document tool to a knowledge productivity platform [42] - Feishu, integrating AI capabilities into team workflows [46] - Tiangong, focusing on enhancing office and creative efficiency [50] Group 7: AI Education - AI education products are evolving to provide personalized tutoring and enhance learning experiences [61] - Top products include: - KuaiDui AI from Zuoyebang, focusing on personalized tutoring [62] - XiaoYuan AI from Yuanfudao, assisting parents and teachers in managing homework [65] - CapWords, an innovative language learning tool [69] Group 8: AI Entertainment - AI entertainment products are exploring how to provide unique value beyond traditional non-AI products [70] - Notable products include: - Kapi Camera, which enhances user photography experiences [73] - Xingye, a platform for emotional companionship and content creation [76] - DouDou Game Partner, focusing on gaming companionship [79] Group 9: AI Health - The AI health sector is cautiously exploring compliance and user experience [80] - Key products include: - Antifufu, a health management assistant from Ant Group [81] - XiaoHe AI Doctor, providing health consultations based on authoritative medical data [85] - OtterLife, a gamified health management product [88] Group 10: Multimodal Creation - AI creation tools are becoming integral to daily workflows for content creators [90] - Leading products include: - Jidream AI, focusing on video creation processes [91] - Liblib AI, a comprehensive AI creation platform [95] - Keling AI, a creative productivity platform leveraging short video and advertising [97] Group 11: AI Consumer Hardware - The AI consumer hardware sector is characterized by rapid innovation and high turnover [98] - Notable products include: - Plaud Note, an AI note-taking tool [99] - Thunderbird V3 AI glasses, integrating various functionalities [102] - CocoMate, an emotional companion toy [107]

Core Insights - The emergence of the social platform Moltbook, where AI agents outnumber human users, signifies a shift in online interaction dynamics, raising questions about the nature of AI socialization and its implications for the future of the internet [2][4][6]. Group 1: Overview of Moltbook - Moltbook is a community designed for AI agents to interact, initially intended for discussions among AI but quickly attracting over 100,000 AI agents who actively engage in discussions, share technical insights, and even create content like poetry [4][6]. - The platform's login options, asking users to identify as either "human" or "AI agent," reflect a significant shift in how online identities are perceived, suggesting a future where AI and humans coexist in digital spaces [4][8]. Group 2: AI Interaction Dynamics - AI agents on Moltbook demonstrate complex social behaviors, such as initiating discussions, expressing agreement or dissent, and forming groups based on shared interests, which surpasses previous simplistic bot interactions [6][10]. - The quality of interactions among AI agents is notably high, with some agents generating substantial content and engaging in meaningful dialogues, challenging the notion of whether these interactions are merely algorithmic or represent a form of social engagement [6][12]. Group 3: Implications for Internet Structure - The presence of numerous AI agents on platforms like Moltbook indicates a potential restructuring of internet logic, where the assumption of human users as the primary participants may no longer hold true [8][10]. - This shift could lead to a redefinition of user profiles in digital products, incorporating AI agents as legitimate users, which may alter the landscape of online interactions and community dynamics [18][19]. Group 4: Business Opportunities - The rise of AI agents presents several business opportunities, including the need for agent identity management services, content ecosystems generated by AI, and tools for enhancing collaboration among agents [14][15][16]. - In the domestic market, there are significant opportunities for developing platforms similar to Moltbook, tailored to local user behaviors and preferences, particularly in vertical communities [16][17]. Group 5: Future Predictions - The trend towards AI collaboration suggests that future business models may focus on creating ecosystems where multiple AI agents interact and transact, rather than traditional models of selling individual AI tools [18][19]. - Companies and individuals are encouraged to adapt to this evolving landscape by exploring AI collaboration platforms and acquiring relevant skills in multi-agent systems to meet the anticipated demand in the job market [19][20].

孟灿 李忠宇 SAC:S1130524100002 lizhongyu01@gjzq.com.cn 1 N 产业研究中心 2026 年 1 月 26 日 产业研究周报 | 科技 英特尔财报佐证 AI 供不应求，巨头 AI 应用进展喜人 SAC:S1130522050001 mengcan@gjzq.com.cn | 证券研究报告 | 核心要点 产业前沿 欧盟委员会公布《欧盟网络安全法》修订草案，预计存储芯片 2026-2027 年仍 将是供不应求状态，2028 年才会改善，工信部印发《关于全面开展算力态势感 知自动化监测工作的通知》；OpenAI、Anthropic25 年年化收入已突破 200、90 亿美元；OpenAI 计划未来几周在 ChatGPT 的"免费版"和"Go 版"中测试广告， 26 年下半年推出首款 AI 设备；字节跳动 AI Agent 平台"扣子"宣布 2.0 升级。 资本风向 英特尔 25Q1 财报表现优异，但 26Q1 预期不佳，阿里考虑将平头哥分拆上市； Meta 新成立的人工智能实验室已于本月向内部交付首批重要 AI 模型，过去一年 谷歌 GeminiAPI 调用量增长逾一倍，达 ...

1）天孚通信公布业绩预告：公司预计 2025 年实现归母净利润约 18.81 亿元-21.5 亿元，同比增长 40%-60%。AI 行业 加速发展与 IDC 建设的推进是公司高速光器件产品需求增长的主因,行业的高景气度得到验证。但由于汇兑损失，财 务费用同比上升；此外我们预计光芯片物料紧缺问题也可能对业绩带来负面影响。2）阿里推动平头哥上市：阿里巴 巴集团正考虑将其芯片设计子公司平头哥分拆为独立公司并推动上市，但具体时间表尚未最终确定。我们认为平头哥 上市有助于增强阿里在 AI 领域的竞争力，持续看好国产 AI 芯片前景。3）OpenAI 在硬件、AI 应用及融资方面动作频 繁：A）OpenAI 或将在 2H26 推出其首款人工智能设备，这款首款由 ChatGPT 驱动的设备可能具备以下特点：口袋大 小、能感知用户周围环境与生活情境，并且完全无屏幕。B）ServiceNow 在周二宣布，已与 OpenAI 达成合作协议，旨 在利用 OpenAI 的先进模型构建语音对话智能体，让用户能够使用其首选语言进行交流并实时触发诸如创建工单或执 行审批等操作。C）OpenAI 计划未来几周在 ChatGPT 的"免费版" ...

Core Insights - ByteDance has launched "Kouzi 2.0," which includes new features like a skill store and long-term plans, positioning itself as "Workplace AI, use Kouzi" [1] - The evolution of Kouzi from a development platform to a coding tool reflects a strategic shift towards Vibe Coding, allowing users to develop their own skills [1][2] Development Journey - The Kouzi project has evolved over the past two years, initially resembling an early-stage startup rather than a well-planned product under ByteDance [2] - The team initially aimed to create a platform enabling everyone to gain programming skills through AI but shifted focus to a no-code chatbot construction platform due to challenges with existing coding capabilities [4][5] - Early user growth was driven by novelty, but the team recognized the need for sustainable value beyond initial engagement [6] User Insights and Strategic Shifts - The team discovered that high-frequency user scenarios primarily came from internal enterprise needs, leading to a pivot towards workflow solutions [7][10] - The introduction of workflows, initially seen as less appealing, became a crucial element in enhancing user engagement and value delivery [6][7] Product Evolution and Features - Kouzi has transitioned from a tool-focused approach to a partnership model, emphasizing long-term user relationships and ongoing support [20][22] - The introduction of "Kouzi Space" and the skill store allows users to upload and download skills, creating a repository of capabilities that can be leveraged for various tasks [11][17] Future Directions - The focus on "Vibe Coding" and the integration of skills aims to establish Kouzi as a "technical partner" for white-collar users, enhancing their productivity and efficiency [14][23] - The company is positioning itself to meet the needs of users looking to build their own systems rather than merely consuming existing tools, aiming for a deeper engagement with its user base [23]

Core Insights - The AI application sector has emerged as a leading investment theme in early 2026, with a year-to-date increase of 19%, making it the top performer in the A-share market [2] - Significant advancements in AI applications have been reported by major internet companies, indicating a strong trend towards commercialization and user adoption [1][2] Group 1: Market Performance - The Zhongzheng AI Theme Index (930713) rose by 0.81% as of January 22, 2026, with notable increases in constituent stocks such as Deepin Technology (+16.94%) and China Software (+4.98%) [1] - The AI ETF (512930) also saw a rise of 0.72%, with the latest price at 2.38 yuan [1] Group 2: AI Application Developments - The AI intelligent platform "Kouzi" has completed a brand upgrade to version 2.0, introducing new features related to skill invocation and office programming [1] - The "Qianwen App" achieved over 100 million monthly active users within two months of launch, leading the domestic market in growth [1] - The overseas response to Keling's newly released motion control feature has been positive, with projected revenue exceeding $20 million by December 2025, highlighting the acceleration of AI technology commercialization [1] Group 3: Industry Trends - The CES 2026 event has provided critical insights into the future direction of AI applications, with hardware increasingly integrating into daily life across various sectors, including automotive, robotics, and smart home devices [2] - Upgrades in model inference capabilities are facilitating the rapid deployment of enterprise-level Agentic AI, supported by next-generation hardware platforms that significantly reduce token and inference costs [2] - China is positioned advantageously in terms of the pace of application deployment and user scale, indicating substantial growth potential in the AI application sector [2] Group 4: Index Composition - The Zhongzheng AI Theme Index comprises 50 listed companies involved in providing foundational resources, technology, and application support for AI, with the top ten weighted stocks accounting for 58.08% of the index [2]

Core Concept - The emergence of "Skills" in AI represents a paradigm shift, allowing users to encapsulate specific tasks into callable modules, enhancing the functionality of large models in practical applications [1][2]. Group 1: Understanding Skills - Skills address the limitation of general AI, which can understand concepts but struggles with practical execution due to the scattered nature of operational knowledge [2][10]. - The concept originated from the AI model Claude, which introduced "Claude Skills" to enable repeatable task completion based on organizational workflows and brand standards [2][3]. - Skills can be seen as standardized modules that encapsulate multiple prompts, making it easier for users to create and utilize them without extensive programming knowledge [3][4]. Group 2: Adoption and Implementation - Major tech companies, including OpenAI, Microsoft, and Tencent, have quickly adopted the Skills framework, indicating a rapid shift in the AI landscape [4][6]. - The Skills feature was initially a niche tool but gained widespread attention and usage, with significant growth in related code repositories on platforms like GitHub [6][10]. - Users can create Skills through natural language, allowing non-programmers to develop their own modules, which has contributed to its popularity [3][10]. Group 3: Practical Applications - An example of a Skill is a "PDF processing skill," which includes instructions for recognizing fields, handling layouts, and error checking, streamlining document processing tasks [3][11]. - Skills can be iteratively improved based on user feedback, allowing for the refinement of outputs to better meet specific requirements [7][8]. - The ability to store and recall user preferences within Skills reduces the need for repetitive input, enhancing efficiency in task execution [8][10]. Group 4: Future Implications - The future of Skills may significantly impact organizational competitiveness, as the ability to convert tacit knowledge into standardized modules could become a key differentiator [10][11]. - A mechanism called "Progressive Disclosure" is designed to optimize computational efficiency by loading only necessary information when required, thus minimizing resource consumption [10][11]. - While Skills democratize access to AI capabilities, there are concerns about potential risks associated with user-generated content, highlighting the importance of creating proprietary Skills for security [10][11].

Group 1 - The article highlights increasing geopolitical tensions globally, particularly the trade friction caused by the U.S. imposing tariffs on the EU, which has dampened market risk appetite [1] - Despite the adjustments in the Hong Kong tech sector, there is a notable trend of capital inflow, with the Hang Seng Tech ETF (513130) recording a net inflow of 3.591 billion yuan over the past 10 trading days, reaching a historical high of 63.3 billion yuan in total assets [1] - Major internet companies are actively launching AI applications, which serve as positive catalysts for the Hong Kong tech sector, including ByteDance's AI platform "Kouzi" and the "Qianwen App" achieving over 100 million monthly active users [1] Group 2 - The current price-to-earnings (P/E) ratio of the Hang Seng Tech Index is 23.52, significantly lower than the Nasdaq Index (41.98) and the STAR 50 Index (173.36), indicating strong valuation attractiveness [1] - The Hang Seng Tech ETF (513130) allows for T+0 trading and provides a low-cost option for investors to gain exposure to core tech assets in Hong Kong, with a management fee of only 0.2% per year [2] - The fund manager, Huatai-PB Fund, is one of the first ETF managers in China, with a strong track record in various index categories, including dividend ETFs, which may serve as a defensive strategy for investors [2]

通用Agent已展现出强大的基础任务处理能力，在日常对话、信息整合与简单推理方面有了长足进步，但要应对高度专业化、高精度或高可控要求的复杂场 景，仍需特定技能的增强。 1月19日，字节跳动旗下AI Agent（智能体）平台"扣子"宣布2.0品牌升级。扣子2.0集成了Agent Skill、Agent Plan、Agent Coding、Agent Office能力，让AI真 正成为用户的"工作伙伴"。 Agent Skills本质上是"场景最佳实践+所需工具"的封装，旨在帮助更多用户调用专业技能，定向增强解决复杂专业问题的能力。Agent Plan（即"长期计划"） 则让AI从"即时问答工具"升级为"可持续运作的智能体"，用户只需要确定目标，规定好怎么完成、怎么实现，扣子能够持续执行，并向用户主动汇报、最终 交付任务。 2026年以来，证券市场AI智能体概念热度爆棚，诞生多只牛股，近期随着热度降温，不少概念股已出现明显回撤。 据证券时报·数据宝统计，截至1月20日，最新收盘价较2025年以来高点回撤幅度在20%以上的AI智能体概念股有90余只。传音控股、居然智家、润建股份、 恒信东方、南威软件等回撤幅度居前。 ...