2026年CES即将开幕，新一轮AI浪潮即将掀起，属于中国企业的AI+硬件出海之门持续敞开。36氪将携手同创伟业和创业观察栏目「摊牌」带您直击2026 CES现场。 图：haivivi创始人李勇，同创伟业投资人王晶，云鲸创始人张峻彬 "新一代的硬件创业公司，从踏上创业的第一天起，脑子里就装着一张世界地图。如果连做高溢价品牌的野心都没有，那就太没想象力了"，如今这句话正成 为越来越多科技、硬件创业者的共识。 全球视野是新一代硬件创业者的共同特质。背靠全球完整、高效的工业体系和完善的供应链网络，新一代具备品牌影响力的中国硬件公司在远洋搏击的战场 上捷报频传。 根据海关总署发布的数据，2025年前三季度我国高技术产品出口额达到3.75万亿，同比增长11.9%。 对中国企业来说，CES几乎成为出海亮相的第一站。越来越多智能硬件公司、AI 创业团队，都把 CES 视为进入全球市场的"首堂必修课"。这场"科技界的春 晚"，也是科技企业开年的最佳营销场域。 AI是近几年CES的核心议题，CES也是观察AI行业进化迭代的窗口。摩根士丹利预判，2026年将成为AI科技硬件迎来爆发式增长的关键之年。而这场爆发的 前哨战，将在20 ...

此次开源意味着硬件厂商、手机厂商和开发者均可基于AutoGLM，在自己的设备或系统中复现一个 能"看懂"屏幕、并模拟真人进行点击、输入、滑动的AI助手。目前，AutoGLM已支持微信、淘宝、抖 音、美团等超过50个高频中文应用的核心场景，其自动化操作能力与此前引发热议的"豆包手机"演示相 似。 智谱深夜开源其核心AI Agent模型AutoGLM。该模型被业界视为全球首个具备"Phone Use"(手机操作)能 力的AI Agent，能够稳定完成外卖点单、机票预订等长达数十步的复杂操作流程。 AutoGLM的开源将大幅降低AI手机的技术门槛，推动AI手机生态从封闭走向开放共创。同时，项目支 持本地与云端部署，确保数据与隐私控制权始终掌握在使用者手中。 ...

Core Insights - The core focus of the news is the announcement by Zhipu that it has open-sourced its core AI Agent model, AutoGLM, which aims to enable AI to perform mobile operations effectively, marking a significant advancement in AI capabilities [1][2]. Group 1: AutoGLM Development and Features - AutoGLM is designed to help users perform tasks on mobile applications, such as placing orders on food delivery apps and managing notifications [1]. - The model is set to be released on October 25, 2024, and is recognized as the first AI Agent with "Phone Use" capabilities [1]. - In 2025, Zhipu plans to launch AutoGLM 2.0, which will utilize algorithms like MobileRL, ComputerRL, and AgentRL to enhance the model's accuracy and generalization by conducting reinforcement learning across thousands of virtual device environments [1]. Group 2: Industry Context and Privacy Concerns - Zhipu is a leading AI large model company incubated by Tsinghua University, with a focus on developing next-generation cognitive intelligence models [2]. - The industry is shifting towards more practical AI agents that can not only communicate with users but also take actionable steps based on user input [2]. - Privacy and network security concerns are significant, particularly regarding the need for user authorization for screen reading permissions, which raises fears about personal information leakage and potential misuse by malicious entities [2][3]. Group 3: Competitive Landscape and Challenges - Recently, Doubao announced a technical preview of its mobile assistant, which has garnered market attention and led to a surge in ZTE's stock price [2]. - Users reported issues with the Doubao mobile assistant, including app crashes and login failures, which may be linked to the need for system-level permissions [3]. - Doubao has stated that while its assistant requires screen reading capabilities, it does not store user screen content in the cloud, aiming to ensure user privacy [3]. - Zhipu has opted for a cloud-based deployment for AutoGLM to allow for action replay, auditing, and intervention, thereby isolating sensitive data [3].

Core Viewpoint - The open-sourcing of AutoGLM by Zhipu AI marks a significant shift in the mobile industry, enabling the creation of AI agents that can operate smartphones similarly to human users, thus democratizing access to advanced AI technology [4][9][19]. Group 1: Open-Sourcing of AutoGLM - Zhipu AI has officially open-sourced its core AI Agent model, AutoGLM, which has been in development for 32 months and possesses the capability to operate mobile devices [1][4]. - The open-sourcing is seen as a "table-turning" move, allowing hardware manufacturers and developers to create personalized AI assistants that can interact with various applications [4][9]. - AutoGLM supports over 50 high-frequency Chinese applications, enabling automated operations similar to the recently launched "Doubao phone" [4][8]. Group 2: Industry Context and Reactions - The announcement comes after the launch of the "Doubao phone" by ByteDance and Nubia, which caused significant disruption in the tech industry and prompted major companies to implement defensive measures against its capabilities [6][8]. - The "Doubao phone" was able to perform complex tasks across applications, leading to a swift market response and concerns over privacy and data control from major internet companies [8][9]. Group 3: Technical Innovations - AutoGLM's development involved creating a framework that allows AI to understand and interact with mobile interfaces, marking a significant evolution from previous systems [12][15]. - The model utilizes Android Debug Bridge (ADB) commands combined with visual models to perform human-like operations, making it difficult for app developers to implement effective defenses [15][18]. - Zhipu AI has addressed privacy concerns by allowing data to remain on the user's device, thus countering arguments from major companies about potential data breaches [18]. Group 4: Implications for the Mobile Industry - The open-sourcing of AutoGLM is expected to reshape the mobile industry, similar to how autonomous driving has transformed the automotive sector, providing a new competitive edge for hardware manufacturers [20]. - Major internet companies face challenges as users can bypass traditional app interfaces, potentially disrupting their established flow of traffic and revenue [23]. - Individual developers are empowered to create specialized AI agents, leading to a more diverse ecosystem of applications and services [24]. Group 5: Conclusion - The open-sourcing of AutoGLM signifies a fundamental shift in user and developer control over mobile technology, indicating the decline of closed ecosystems and the rise of a new era in intelligent mobile interactions [28].

每经AI快讯，12月9日，智谱宣布开源其核心AI Agent模型 AutoGLM，该模型是一款具有"Phone Use"（手机操作）能力的 AI Agent，能够稳定地完成外卖点 单、机票预订等复杂操作流程。 （文章来源：每日经济新闻） ...

智谱宣布开源其核心AI Agent模型AutoGLM，该模型是一款具有“Phone Use”（手机操作）能力的 AI Agent，能够稳定地完成外卖点单、机票预订等复杂操作流程，操作步骤可长达数十步。智谱表示，希望通过开源和私有化部署，企业和开发者可以在自己的合规环境中完整掌控数据、日志和权限，可以让手机，成为专属自己的AI手机。外汇交易员 (@myfxtrader):在经历微信与银行金融类APP的风控干预后，豆包手机助手宣布对金融类APP操作、激励刷分操作与部分游戏应用的操作能力进行限制。 https://t.co/mu3ZyzpqdM ...

人民财讯12月9日电，12月9日，智谱开源其核心AI Agent模型AutoGLM。该模型是具备"Phone Use"（手机操作）能力的AI Agent，能够稳定完成外卖点单、机票预订等长达数十步的复杂操作流程。 这意味着硬件厂商、手机厂商和开发者均可基于AutoGLM，在自己的设备或系统中复现一个能"看 懂"屏幕、并模拟真人进行点击、输入、滑动的AI助手。目前，AutoGLM已支持微信、淘宝、抖音、美 团等超过50个高频中文应用的核心场景，其自动化操作能力与此前引发热议的"豆包手机"演示相似。 ...

12月9日上午，智谱宣布开源其核心AI Agent模型AutoGLM。该模型是一个具备"Phone Use"（手机操作）能力的AI Agent，能够稳定完成外卖点单、机票预订等长达数十步的复杂操作 流程。 此次开源意味着硬件厂商、手机厂商和开发者均可基于AutoGLM，在自己的设备或系统中复现 一个能"看懂"屏幕、并模拟真人进行点击、输入、滑动的AI助手。目前，AutoGLM已支持微信、 淘宝、抖音、美团等超过50个高频中文应用的核心场景，其自动化操作能力与此前引发热议 的"豆包手机"演示相似。 AutoGLM的开源将大幅降低AI手机的技术门槛，推动AI手机生态从封闭走向开放共创。同时， 项目支持本地与云端部署，确保数据与隐私控制权始终掌握在使用者手中。（转载自新浪科技） ...

Core Insights - The core point of the news is that Zhiyu has announced the open-sourcing of its core AI Agent model, AutoGLM, which is capable of performing complex operations on mobile devices, such as food ordering and flight booking [1][2] Group 1 - The AutoGLM model can execute long and complex operation processes, demonstrating capabilities in "Phone Use" [1][2] - The open-sourcing allows hardware manufacturers, mobile device makers, and developers to replicate an AI assistant that can "understand" screens and simulate human interactions like clicking, inputting, and sliding [2] - AutoGLM currently supports over 50 high-frequency Chinese applications, including WeChat, Taobao, Douyin, and Meituan, showcasing its automation capabilities similar to the previously discussed "Doubao phone" demonstration [2] Group 2 - The open-sourcing of AutoGLM is expected to significantly lower the technical barriers for AI mobile applications, promoting a shift from a closed to an open collaborative AI mobile ecosystem [2] - The project supports both local and cloud deployment, ensuring that data and privacy control remain with the users [2]

Core Insights - The future may see the emergence of "one-person unicorn companies," where a single CEO manages multiple AI agents, while human employees evolve into "goal architects" who can leverage AI to enhance their capabilities and deliver greater value [1][4]. Group 1: Future of Work - Human employees will not be easily replicable, whereas "super intelligent agents" can be infinitely replicated, enabling rapid global expansion for companies [1][4]. - The core competitive advantage for future enterprises will hinge on three key aspects: early adoption and use of AI agents, selection of the most advanced AI agents, and continuous training of these agents using closed-loop data to enhance their capabilities [1][4]. Group 2: Strategic Recommendations - Companies should prioritize early and swift integration of AI agents into their operations [1][4]. - The importance of utilizing closed-loop data for ongoing training of AI agents is emphasized as a critical factor for success [1][4].