Core Insights - OpenAI has launched the "Company Knowledge" feature for ChatGPT, aimed at Business, Enterprise, and Education users, enabling integration of internal information resources for precise business-related responses [1][2] Feature Overview - The "Company Knowledge" feature integrates data from popular work applications like Slack, SharePoint, Google Drive, and GitHub, leveraging the new GPT-5 model for efficient information retrieval [2] - It provides clear citations for each response, enhancing the credibility and traceability of the information [2] Practical Applications - The feature can automatically compile information for meetings, saving users time by generating briefs from various sources such as Slack messages and customer emails [2] - It assists in transforming customer feedback into strategic documents and summarizing project performance, addressing a wide range of enterprise needs [2] User Experience - Users can easily activate the feature and connect relevant applications, with real-time display of the information retrieval process and sources [3] - ChatGPT will only access data that users are authorized to view, ensuring compliance with existing permission systems [3] Privacy and Security - OpenAI has implemented measures to protect enterprise data, including not using company data for model training and employing industry-standard encryption [3] - Features like SSO and SCIM for access management, along with IP whitelisting, enhance security [3] Limitations and Future Plans - Users must manually enable "Company Knowledge" for each new conversation, and the feature currently lacks internet search capabilities and the ability to generate charts [4] - OpenAI plans to integrate "Company Knowledge" with all ChatGPT functionalities and expand support to more enterprise tools in the future [4] Availability - The "Company Knowledge" feature is now available to all enterprise, education, and business users since its announcement [5]

在美欧关税争端持续升级的背景下，新浪财经社区凭借其独特的信息整合能力和实时互动机 制，成为用户获取动态、参与讨论的核心平台。其高讨论度与用户黏性源于以下三方面： 1 ．快而不躁： 实时信息筛选与风险预警 面对关税政策的频繁变动，新浪财经社区通过实时弹幕过滤系统，自动提取关键词（如" 200% 关税""欧盟反制"），第一时间将核心信息置顶。例如， 2025 年 3 月特朗普威胁 对欧盟酒类加征关税的消息，社区弹幕迅速聚焦"烈酒出口""供应链中断"等关键点，帮助 用户跳过冗长新闻，直达核心矛盾。 此外，平台利用多空情绪热力图，量化用户讨论中的看涨 / 看跌倾向。例如，当欧盟宣布 对美汽车加征关税时，热力图显示"看空"情绪占比骤升，系统随即推送历史关税战案例 （如 2018 年美欧钢铝关税冲击），辅助用户理性判断市场风险。 2 ．准而有据： 数据交叉验证与产业一线联动 新浪财经社区构建了产业数据众筹机制，吸引钢铁贸易商、汽车制造商等从业者分享一线数 据，这类数据与官方统计形成互补，为用户提供多维视角。同时，平台整合 20 余家期货公 司研报，标注"基差修复""库存周期"等实战关键词。例如，围绕苹果减产展开专题讨论， ...