以下文章来源于财联社AI daily ，作者周子意 而最新的报道称， 英伟达首席执行官黄仁勋近期私下向业内同仁强调，最初的1000亿美元协议并非具有约束力，也尚未最终确定。 财联社AI daily . 财联社及科创板日报旗下产品——未来已来，AI前沿，独家、深度、专业！ 据报道， 英伟达和OpenAI之间价值1000亿美元的合作交易谈判已陷入停滞。 去年9月英伟达曾宣布，计划向OpenAI投资高达1000亿美元，以帮助其训练和运行最新的人工智能模型。但据知情人士最新透露， 这一计 划因英伟达内部人士对这笔交易存疑而停滞不前。 据媒体援引知情人士的话报道，这两家公司目前正在重新考虑其合作关系的未来，最新讨论包括在OpenAI当前融资轮中进行数百亿美元的 股权投资。 这项交易的搁置显然对OpenAI构成了不小的打击。这笔交易本可以为OpenAI提供所需的资金和渠道，使其能够购买关键的先进芯片，以在 日益激烈的竞争环境中保持领先地位。 并且值得一提的是，这家ChatGPT制造商正为2026年底前上市做准备，过去一年一直在争取获得大量算力以支撑未来产品和增长。 竞争担忧 根据两家公司于去年9月签署的谅解备忘录，英伟达 ...

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月24日，摩根大通在最新研报中称，1月16日，OpenAI宣布将在未来几周内开始在ChatGPT的免费版和 Go版本（轻量级订阅版）中测试广告，这标志着其变现策略发生了实质性转变。 研报称，这一转向并非偶然—— 面对接近10亿的周活跃用户中95%为免费用户的现状，以及未来八年高 达1.4万亿美元的基础设施投入承诺，广告收入已从可选项变为商业必需。 摩根大通分析指出， ChatGPT的广告模式将有别于传统社交平台的"注意力经济"，转而通过捕捉用户的 高商业意图来获取更高的千次展示费用（eCPM） 。目前ChatGPT约12%的查询具有搜索性质，且用户购 买意向强烈。 硬·AI 作者 | 董 静 编辑 | 硬 AI 正式告别"免费午餐"时代，OpenAI终于按下了广告变现的按钮。摩根大通认为，随着ChatGPT宣布将在未 来几周内开始测试广告，这家AI巨头正式迈出了从单纯依赖订阅模式向多元化商业变现转型的关键一步。 商业化"三驾马车"驱动 广告是必然选择 摩根大通认为，OpenAI部署广告是基于三个关键因素的商业必然： 首先是庞大的免费用户基数 摩根大通称，OpenAI面对95%的免费用户占比及未来八年1.4万 ...

正式告别"免费午餐"时代，OpenAI终于按下了广告变现的按钮。摩根大通认为，随着ChatGPT宣布将在未来几周内开始测试广告，这家AI 巨头正式迈出了从单纯依赖订阅模式向多元化商业变现转型的关键一步。 1月24日，摩根大通在最新研报中称，1月16日，OpenAI宣布将在未来几周内开始在ChatGPT的免费版和Go版本（轻量级订阅版）中测试 广告，这标志着其变现策略发生了实质性转变。 研报称，这一转向并非偶然——面对接近10亿的周活跃用户中95%为免费用户的现状，以及未来八年高达1.4万亿美元的基础设施投入承 诺，广告收入已从可选项变为商业必需。 摩根大通分析指出，ChatGPT的广告模式将有别于传统社交平台的"注意力经济"，转而通过捕捉用户的高商业意图来获取更高的千次展 示费用（eCPM）。目前ChatGPT约12%的查询具有搜索性质，且用户购买意向强烈。 该行认为，这一策略不仅旨在开辟新的收入流，更意在构建一个涵盖免费、Go（8美元/月）、Plus（20美元/月）及Pro（200美元/月）的 清晰价值阶梯，通过广告体验倒逼用户向付费层级转化。 同时，摩根大通认为，作为首个在消费级应用中测试广告的主流大语言 ...

Core Insights - OpenAI is shifting its strategic focus from consumer products to the enterprise market, aiming to provide comprehensive support for business clients [1][2][3] Group 1: Strategic Shift - OpenAI's notable actions in 2025 were primarily centered around consumer products, but it is now emphasizing enterprise solutions [2] - CEO Sam Altman indicated a desire for OpenAI to become a one-stop shop for all AI needs of businesses, including tools like ChatGPT and Codex [2] - A new initiative is being developed to assist large companies in integrating AI into various business operations, such as customer service and data organization [2][4] Group 2: Competitive Landscape - OpenAI aims to attract enterprise clients from competitors like Anthropic, which has gained attention with its tools Claude Code and Cowork [3][5] - OpenAI's products have not met expectations in certain areas, prompting a need for upgrades to compete effectively [3] - The competitive landscape is intensifying, with Google Gemini emerging as a significant challenger, capturing approximately 20% of the market share [6][8] Group 3: Market Dynamics - OpenAI's market share has decreased from nearly 90% to around 65% over the past year, indicating a loss of dominance [6][8] - Monthly unique visitors for ChatGPT increased by 30.6 million, with a penetration rate rising from 71% to 77.3% [8][9] - Other competitors like Google Gemini and Grok are also experiencing significant growth in user engagement and market penetration [8][9] Group 4: Revenue and Business Model - Currently, about 40% of OpenAI's revenue comes from enterprise clients, with expectations to increase to 50% by the end of the year [4] - OpenAI has restructured its sales approach to streamline the process for enterprise clients, focusing on a unified sales strategy [4]

科技界一条重磅消息"刷屏"了。 苹果和 Google 发布联合声明，称双方已达成合作协议，苹果决定常规性地接入谷歌 Gemini，将其作为苹果 AI 的"核心大脑"。 这意味着，未来 iPhone、iPad 和 Mac 上的 AI 能力，其"智力"来源很大程度上是谷歌的技术。 苹果展示"Siri 请求是否可以向ChatGPT发送询问" 同时，相关协议显示，基于此技术，全新且更具个性化的 Siri 预计将在今年晚些时候推出。 据悉，该模型将负责 Siri 系统级的推理、个性化服务 （如理解屏幕内容、跨App操作） 。这意味着它能像人一样进行多步骤的逻辑思考，处理能力较旧版将有显 著提升。 举个例子，你截图了一张机票信息，现在的 Siri 可能只能识别图上的字；而未来的 Siri 能直接理解你的动机： 这是 xx 的航班， TA 多久到，我可以帮你设置个闹钟，并且查一下去机场的路况。 这种改变将让苹果 AI 从一个"聊胜于无"的玩具，变成一个真正能帮你干活的生产力工具。 值得一提的是，OpenAI 此前曾和苹果达成了合作。2024 年 WWDC上，双方首次宣布联手。 彼时，ChatGPT 是作为 Siri 的"补 ...

Core Viewpoint - The CEO of French AI startup Mistral, Arthur Mense, claims that the notion of China lagging behind the U.S. in AI technology is a "fairy tale" and asserts that China's open-source AI capabilities may pressure U.S. CEOs [1][3] Group 1: Company Insights - Mistral is projected to exceed €1 billion in revenue by the end of this year [1] - The company plans to invest a similar amount in high-performance computing chips and related infrastructure for AI model development and operation [1] - Mistral's valuation reached $13.7 billion during a funding round last year, with Dutch chipmaker ASML as a major investor [1] Group 2: Industry Context - AI is becoming a significant geopolitical force with the potential to reshape economies and labor markets in the coming years, with companies and nations investing tens of billions of dollars in AI infrastructure [1] - The AI market is currently dominated by the U.S. and China, while Europe is seeking differentiation [1] - Many U.S. AI models, such as Google's Gemini and OpenAI's ChatGPT, are closed-source, which can lead to higher costs and less flexibility compared to China's leading position in open-source model development [4]

Core Viewpoint - The CEO of French AI startup Mistral, Arthur Mense, claims that the notion of China lagging behind the U.S. in AI technology is a "fairy tale" and asserts that China's open-source AI capabilities may pressure U.S. CEOs [1][3] Group 1: Company Insights - Mistral is projected to exceed €1 billion in revenue by the end of this year [1] - The company plans to invest a similar amount in high-performance computing chips and related infrastructure for AI model development and operation [1] - Mistral's valuation reached $13.7 billion during a funding round last year, with Dutch chipmaker ASML as a major investor [1] Group 2: Industry Context - AI is becoming a significant geopolitical force with the potential to reshape economies and labor markets in the coming years, with companies and nations investing tens of billions of dollars in AI infrastructure [1] - The AI market is currently dominated by the U.S. and China, while Europe is seeking differentiation [1] - Many U.S. AI models, such as Google's Gemini and OpenAI's ChatGPT, are closed-source, which can lead to higher costs and less flexibility compared to China's leading position in open-source model development [5]

【文/观察者网 齐倩】 在欧洲焦虑于人工智能（AI）落后中美之际，谷歌公司前首席执行官、Alphabet公司前执行董事长埃里 克·施密特在瑞士达沃斯世界经济论坛期间给欧洲上强度。 施密特资料图 彭博社指出，在AI科技竞赛中，欧洲基本被美国和中国甩在身后。法国Mistral AI初创公司在欧洲领 先，去年在一轮融资中估值达到137亿美元，但这与OpenAI超5000亿美元的估值相比只是零头。 据彭博社消息，当地时间1月20日，施密特表示，欧洲必须投资建设自己的开源AI实验室，并解决不断 飙升的能源价格，否则将很快发现自己依赖中国模型。 "在美国，大多数公司正转向闭源，这意味着它们将被购买、授权等，"施密特在论坛上说，相较之 下，"中国企业则主要采用开放权重、开源的方式"。他表示，这意味着，"除非欧洲愿意为欧洲AI模型 投入巨资，否则欧洲最终会使用中国模型"。 施密特随后暗戳戳使坏，称若真如此，"这对欧洲来说可能不是好结果"。 彭博社介绍称，许多美国AI模型，如谷歌Gemini和OpenAI的ChatGPT，都是闭源的，这意味着开发公 司不会公开底层代码供下载或审查。虽然这能为用户提供更流畅、更统一的体验，但通常 ...

Core Insights - The article emphasizes the significant impact of AI on global economic growth and stock market performance, particularly in the U.S. and China, highlighting the importance of AI investments in driving GDP growth and stock returns [1][11]. Economic Contribution - U.S. technology hardware and software investments are projected to contribute 0.8 percentage points to the annualized GDP growth of 2.5% in the first three quarters of 2025 [2]. - Since 2023, U.S. non-farm business sector labor productivity has increased by 7.2% [4]. Stock Market Performance - Since the launch of ChatGPT at the end of 2022, the "Magnificent Seven" stocks in the U.S. contributed 45 percentage points to the S&P 500 index returns, accounting for over half of the total returns [1]. - In Hong Kong, seven leading tech stocks contributed 14 percentage points to the Hang Seng Index returns, representing 40% of the total returns since the launch of DeepSeek in early 2025 [7]. AI Industry Dynamics - The U.S. has a first-mover advantage in AI infrastructure, models, and talent, while China is rapidly catching up, particularly in open-source models [16][19]. - The global cloud computing market is expected to reach $692.9 billion in 2024, with North America holding a 54.3% market share and China at 16.8% [20]. Investment Landscape - U.S. nominal investment in AI technology hardware and software is approximately $1.05 trillion, accounting for 3.4% of nominal GDP, while China's investment is around $650 billion, also about 3.3% of nominal GDP [42]. - The AI investment landscape shows that the U.S. is primarily driven by private sector investments, while China's investments are significantly supported by government funding [52]. Funding Sources - U.S. AI investments are predominantly led by the private sector, with a total investment of $552 billion, compared to China's $900 billion, which includes substantial government support [52]. - The Chinese government has invested approximately $750 billion in AI, significantly higher than the U.S. government's $110 billion [57]. Investment Focus - U.S. investments are heavily focused on data centers and supporting infrastructure, while China is investing more in chip development and AI models [58]. - In 2025, U.S. leading companies are expected to invest $4 billion in the foundational layer, with 88% directed towards infrastructure [59]. Talent and Research - The number of AI researchers in China has grown significantly, with Chinese researchers holding three times as many AI patents as their U.S. counterparts [41]. - The rapid increase in AI talent in China is enhancing its research capabilities, narrowing the gap with the U.S. [41]. Future Outlook - The article suggests that the U.S. will continue to invest heavily in data centers and energy equipment, which will drive demand for related supply chains in China [88]. - China's ongoing need for chips and AI infrastructure indicates a sustained demand and policy support in the semiconductor sector [88].