本網站使用瀏覽器紀錄 (Cookies) 來提供您最好的使用體驗，我們使用的 Cookie 也包括了第三方 Cookie。相關資訊請訪問我們的隱私權與 Cookie 政策。如果您選擇繼續瀏覽或關閉這個提示，便表示您已接受我們的網站使用條款。 關閉
全球 AI CRM 領導者 Salesforce 今日宣布 Slackbot 正式全面上市。此項發布不僅是產品功能的升級，更是 Salesforce 邁向「代理式企業（Agentic Enterprise）」的重要里程碑。Salesforce 將 Slackbot 定位為內建於 Slack 的「個人智慧代理（Personal Agent）」，旨在重新定義人類與 AI 的協作模式。Slack 使用者無需繁瑣設定，即可透過 Slackbot 將 AI 代理應用融入既有的工作流程。 深度整合企業資料 實現專業 AI 代理應用 Salesforce 致力於透過平台整合全面釋放 AI 的價值，而 Slackbot 的推出，正是為了解決企業 AI 應用落地的「最後一哩路」。它將 Slack 轉化為「代理式工作作業系統（Agentic Work OS）」，讓 AI 不再只是被動回應的對話助理，而是能理解使用者工作流程、並主動執行任務的專業代理。透過串接企業資料庫、Salesforce CRM 資料與第三方系統，Slackbot 能提供具備深度情境理解的智慧服務，進一步優化人類與 AI 的協作方式。 Salesforce 共同創辦人暨 Slack 技術長 Parker Harris 表示：「Slackbot 不僅是 AI 輔助工具，更是協助企業進入『代理式企業』的關鍵入口。透過將 AI 深度整合至包含企業數據、工作流程與對話的 Slack 平台中，我們期望能以企業級 AI 賦能每一位員工，提升整體組織生產力。」 五大關鍵技術 跨系統無痛整合工作流程 全新的 Slackbot 具備多項企業級關鍵能力，能有效消除跨系統操作的隔閡，在提升工作效率的同時，也透過 Salesforce 的技術基礎，確保企業資訊安全無虞： 無需訓練的個人化代理：Slackbot 深度內建於 Slack 平台，無需額外模型訓練或繁複設定，即可理解對話紀錄、檔案內容與團隊協作脈絡，提供貼近使用者實際工作情境的精準回應，而非僅止於一般搜尋結果。 深度整合 Salesforce 客戶數據：透過與 Salesforce CRM 的無縫串接，Slackbot 能在會議前自動彙整客戶背景、年度經常性收入（ARR）與歷史互動紀錄，並結合 Slack 內部討論內容，協助團隊做出更具策略性的判斷。 從對話直接驅動成果：員工只需以自然語言與 Slackbot 對話，即可快速產出會議摘要、撰寫專案簡報，並完成會議安排、提醒設定等行政工作，讓流程自動化真正融入日常工作。 開放式架構與多模態協作：Slackbot 採用開放式架構，除支援 Salesforce Agentforce，也能與包括 OpenAI 的 ChatGPT、Anthropic 的 Claude 等第三方 AI 代理協同運作，成為企業內部整合跨系統工作的 AI 指揮中樞，處理跨系統的複雜工作流。 企業級信任與資安合規：Slackbot 建立於 Salesforce 的資安信任層（Trust Layer）之上，嚴格落實權限控管與存取規範，確保企業機敏資料能在安全、可控的環境中被運用。 Slackbot 即日起開放 Business+ 與 Enterprise+ 用戶使用，並預計於 2 月底全面開放予全球使用者（包含台灣）。Salesforce 也將持續擴展 AI 代理生態系，致力打造人類與 AI 高效協作的全新工作環境。
Actian AI Analyst 專為實現企業級準確度而設，運用語義知識圖譜，提供必要的業務情境，從根源解決 AI 幻覺問題 德州圓石城2026年3月11日 /美通社/ -- HCLSoftware 旗下數據及人工智能 (AI) 部門 Actian 今天發表 Actian AI Analyst（前稱 Wobby）。此對話式分析方案配備 Steward Agent，能以智能方式建立並持續更新語義層，協助企業更快獲取價值，並提供可靠且可供審核的洞察見解。 Actian AI Analyst 結合 AI 分析與確定性語義，讓企業能擴展分析存取、消除商業智能瓶頸，並有信心地按見解果斷行動。 基礎 AI 模型在簡單測試中往往極度精準，但面對複雜的企業數據，卻容易產生幻覺及邏輯偏差。 事實上，Stanford HAI 發佈的《2025 年 AI 指數報告》(2025 AI Index Report) 指出，個別模型在簡單規劃任務上準確率高達 97.8%，但隨著推理越來越複雜，其準確率便會斷崖式下跌，部分模型更幾乎跌至零。1這些錯誤常常無法追溯，導致嚴重的信任危機。 Actian 產品總監 Guillaume Bodet 表示：「現今企業紛紛將 AI 融入分析流程，務求加快產生洞察見解及決策；但若空有速度而缺乏信任，終究是弊大於利。 有別於容易出錯的文字轉 SQL 工具，Actian AI Analyst 建基於受管治的企業數據，並以全面追溯為設計核心，讓企業能安全地為業務用戶普及分析功能，準確度及可信性絲毫不減。」 Steward Agent：智能構建數據基礎 全新 Steward Agent 能根據現有數據倉庫架構及業務文件，以智能方式構建並維護受管治的數據根基，從而加快 AI 部署，告別繁複的人手作業。 語義建模：將現有的業務詞彙表及文件轉化為語義知識圖譜，讓企業在幾天至幾週內，便能從原始數據躍升至可供查詢的洞察見解。 一致的業務邏輯：制定並維護共享的業務邏輯，確保跨部門的定義一致，避免指標有所偏差。 主動邏輯維護：自動偵測並更新業務定義的變動，讓語義層與時並進，不必人手重新配置。 Actian AI Analyst：精準分析 規模擴展 Actian AI Analyst 提供可信賴的對話式分析，讓業務用戶透過日常語言探索精細數據、發掘洞察見解。 方案將所有查詢都規範在受管治的語義層內，讓團隊根據已核實的資訊，即時作出反應。 對話式介面：以自然語言介面取代靜態控制面板，業務用戶可隨時追問，即時獲得分析，完全不用撰寫 SQL 語句。 約束執行引擎：採用規則為本的方法，確保每個答案均建基於企業的特定邏輯，杜絕因任意生成 SQL 而導致的 AI 幻覺。 全面追溯：清楚呈現每個答案背後的推理步驟、篩選條件及計算過程，讓用戶輕鬆核實結果，快速解決數據不一致的問題。 無縫整合與行動 Actian AI Analyst 直接整合企業現有工作流程，大大縮短從提問到制定策略決定的時間。 Slack 與 Microsoft Teams 整合：將對話式分析功能直接帶入現有的協作工具，讓用戶查詢、獲取數據和分享結果時無須來回切換應用程式。 對話上下文串聯：保留追問過程中的完整對話脈絡，讓分析可以層層深入，洞察見解自然演化，不必每次重新開始。 簡化營運：讓業務用戶自助獲取準確答案，減輕數據團隊的營運負擔。 鋼絲轉化與塗層技術材料科學領導企業 Bekaert 的 AU 主管 Pieter Seyns 表示：「Actian AI Analyst 將我們的商業智能 (BI) 轉化為對話，而非開發週期。 洞察見解更快，控制面板更少——Actian AI Analyst 就是如此簡單實用。」 若然機構希望將這些功能推廣至全企業，可結合 Actian AI Analyst 與 Actian Data Intelligence Platform，構建閉環系統，讓受管治的數據定義直接融入即時的對話式分析之中。 供應情況 Actian AI Analyst 及 Steward Agent 現已推出。 未來幾個季度，這些智能代理將陸續整合至 Actian 的整體產品陣容。 如欲深入了解，歡迎查閱網誌或瀏覽Actian AI Analyst 網頁。 關於 ActianActian 讓企業能充滿信心地大規模管理及管治數據。 Actian 數據管理及數據智能解決方案有助簡化複雜的數據環境，並加快提供可供使用的 AI 數據。 Actian 方案以靈活見稱，能在本地部署、雲端及混合環境中無縫整合，表現穩定可靠。 如欲進一步了解 HCLSoftware 旗下數據與 AI 部門 Actian，歡迎瀏覽 actian.com。 1 Stanford University Human-Centered Artificial Intelligence (HAI)，《2025 年 AI 指數報告》，2025 年 4 月。 傳媒聯絡人：Jennifer Harbour公關顧問Jennifer.Harbour@actian.com Danielle Lee副總裁 – 全球分析師關係與公關Danielle.Lee@actian.com
New innovations include expanding AI Companion 3.0 more broadly across the Zoom platform and introducing custom AI agents to orchestrate workflows in Zoom and third-party systems Zoom’s system of action drives conversation to completion with new agentic workflows and CX AI features across Zoom Workplace, Phone, and CX products LAS VEGAS, March 10, 2026 (GLOBE NEWSWIRE) -- Zoom Communications, Inc. (NASDAQ: ZM) today announced the expansion of its enterprise agentic AI platform, introducing new workflow orchestration capabilities across Zoom Workplace, Zoom Phone, and Zoom CX. The updates extend Zoom’s system of action for modern work, enabling organizations to automate tasks, trigger cross-system workflows, and turn meetings, calls, and customer interactions into completed business outcomes. As enterprises scale artificial intelligence initiatives, many remain limited to AI assistants that summarize conversations but rely on manual follow-through across disconnected systems. Moreover, while collaboration tools capture dialogue and systems of record store data, execution often remains fragmented. Zoom’s agentic AI platform is designed to close this gap by embedding workflow automation directly into meetings, calls, chat, and contact center interactions so conversations automatically trigger action across enterprise systems. “The next phase of enterprise AI will be defined by the ability to move from conversation to action,” said Velchamy Sankarlingam, president of Product & Engineering at Zoom. “Zoom’s agentic AI platform is designed to orchestrate action across systems, turning every meeting, call, and customer interaction into a trigger for workflow automation.” Zoom enterprise agentic AI 3.0 platform: Key enhancements The enhancements announced today include: Custom and prebuilt AI agents with no-code orchestration New third-party integrations for AI Companion, including 10 secure enterprise search connectors New AI-first canvases (Zoom AI Docs, AI Sheets, and AI Slides) Zoom Phone Mobile SMS for Zoom Virtual Agent AI Receptionist AI Expert Assist 3.0 for Zoom Contact Center Natural-language customer workflow orchestration Meeting security enhancements AI Companion: Premium AI built for businesses The Zoom AI Companion offering is significantly expanding, with AI Companion 3.0 rolling out across the Zoom Workplace app, Zoom Business Services, and Workvivo, to accelerate conversation-to-completion workflows via the 3.0 AI tech stack. Previously available only in a web browser, this powerful upgrade leverages Zoom’s expanded federated AI platform to securely connect conversations, enterprise data, and apps, and drive meaningful outcomes for organizations. The platform’s success is highlighted by the rapid growth of AI Companion Monthly Active Users, having more than tripled year-over-year in Q4 FY26. Zoom also introduced new capabilities for its Custom AI Companion add-on, empowering organizations to build custom AI agents that act on their behalf — streamlining activities that include retrieving insights, automating tasks, and orchestrating workflows across third-party systems such as Salesforce, Slack, and ServiceNow. New Custom AI Companion capabilities include: Build custom agents and workflows: Organizations can save time on repetitive tasks by leveraging prebuilt agents for Sales, IT, and Marketing, and deploy their own custom AI agents and workflows — with no coding required. Connect additional enterprise data: New third-party integrations for personal workflows and agent builder, along with 10 new secure connectors for intelligent retrieval, enable users to synthesize enterprise data across applications to derive actionable insights from multiple data sources, integrating with third-party platforms such as Salesforce, ServiceNow, Box, Google Drive, and OneDrive. Personalization and memory: AI Companion learns from user context, such as their role, preferences, and focus areas, to deliver tailored insights and recommendations. Teams will be able to easily convert meeting conversations and insights into structured documents, data analysis, and presentation content with the expansion of new AI canvases, Zoom AI Docs, AI Sheets, and AI Slides, which will work seamlessly within Zoom Meetings for real-time co-creation without switching tools. Zoom also announced Zoom AI Services, a new suite of enterprise‑grade AI APIs for speech, language, reasoning, and more. These services enable developers to leverage the same transcription, translation, summarization, deep reasoning, and image‑processing technologies that power Zoom’s own products. Zoom Workplace: Simplify and unify the workday The next generation of Zoom Workplace, an all-in-one work platform, introduces agentic AI and a simplified interface. New features and UX updates include: UX simplification: A cleaner, more consistent interface across desktop, mobile, and web, reducing friction and cognitive load. AI Companion will have its own tab, providing a central, conversational work surface for real-time queries, writing, and workflow automation. Zoomie group assistant: A new team facilitator that works across Zoom Rooms, Meetings, and Chat, designed to drive better meeting outcomes by keeping teams aligned, on track, and focused on core objectives. AI-first Chat: Custom agents available for use in Chat, topic-based summaries, and “For You” prioritization to surface what matters most. Live voice translator: Break down language barriers with real-time audio translation for Zoom Meetings, so participants can speak in their native language while others hear the translation in their preferred language. (Available at launch in five languages, with more coming soon.) Deepfake risk detection: A new advanced meeting security technology intelligently detects synthetic audio or video in meetings and provides real-time alerts. Zoom Phone: Transform conversations into action To enhance its offering for Zoom Phone, which supports more than 10 million seats globally, Zoom announced new capabilities that prioritize a mobile-first experience and use AI to deliver better customer engagements and automate post-call follow-ups: Agentic workflows: Automatically execute tasks such as drafting emails or sending out summaries. SMS for Zoom Virtual Agent AI Receptionist (formerly AI Concierge): New SMS capabilities will enable the 24/7 virtual receptionist to handle customer engagements via text, answer questions, collect information, support scheduling flows, and escalate to a human when needed. Zoom Phone Mobile: A new, intuitive mobile experience lets users make business calls from their native dialer while retaining AI calling capabilities wherever work happens. Customer Engagement Pack (formerly Power Pack): Expanded analytics and AI-driven insights to help customer‑facing teams stay responsive by prioritizing high‑value calls, enhancing queue visibility, and tracking performance across calls and SMS. As a result, Zoom Phone now delivers intelligent call summaries, automated follow-ups, and real-time insights across devices. Zoom CX and Revenue Workflows: AI-Powered Contact Center and Sales Automation Zoom CX and Zoom Revenue Accelerator (ZRA) will soon introduce expanded agentic automation for contact center and revenue workflows, including Zoom Virtual Agent 3.0, to improve resolution speed and customer outcomes. AI Expert Assist 3.0: A real-time, agentic AI layer for the contact center that acts as an active collaborator, empowering agents and supervisors with contextual guidance, intelligent orchestration, and automated task execution for faster resolutions and measurable operational efficiency. CX Insights: Helps business and CX leaders ask questions about customer experience, workforce, and automation performance in natural language and receive clear, reasoned insights by analyzing interaction transcripts, operational signals, and CX metrics across a shared context. Customer Workflow Orchestration: Enables organizations to design and automate customer journeys across systems, channels, and teams using natural-language workflow creation triggered by contact center events or signals from enterprise systems, such as CRM or ERP. Advanced Quality Management for Zoom Virtual Agent: Provides CX leaders with deeper visibility into automated interactions, performance trends, and resolution outcomes across their entire workforce, both human and virtual agents. Zoom Revenue Accelerator enhancements: New AI Sales Assist delivers real-time coaching during sales conversations by surfacing prompts, competitive insights, and next best actions directly within the call flow. Ask ZRA introduces conversational intelligence for revenue leaders, enabling natural language analysis across calls and deals. Zoom Customer Experience continues to see accelerating, high-double-digit growth, with paid AI included in each of Zoom’s top ten CX deals this past quarter and growing competitive displacement momentum. To learn more about Zoom’s announcements at Enterprise Connect, including availability of the new features, visit the Zoom newsroom and stop by Zoom’s booth (#508) for demos of these innovations. About ZoomZoom (NASDAQ:ZM) is a system of action for modern work, turning live collaboration into completed results. From entrepreneurs to global enterprises, customers choose Zoom to seamlessly collaborate, communicate, and drive outcomes across meetings, phone, contact center, and more — all with the built-in assistance of Zoom AI Companion. Founded in 2011, Zoom is headquartered in San Jose, CA. For more information, visit zoom.com. Zoom Public RelationsLacretia Nicholspress@zoom.us
Designed to provide enterprise-grade accuracy, Actian AI Analyst uses a Semantic Knowledge Graph to provide business context needed to address AI hallucinations ROUND ROCK, Texas, March 10, 2026 /PRNewswire/ -- Actian, the data and AI division of HCLSoftware, today introduced the Actian AI Analyst (formerly Wobby), a conversational analytics solution featuring the Steward Agent that intelligently generates and provides ongoing maintenance of the semantic layer to speed time to value and deliver trusted, auditable insights. By combining AI-driven analysis with deterministic semantics, Actian AI Analyst allows organizations to scale analytics access, reduce business intelligence bottlenecks, and act on insights with confidence. Basic AI models generally claim high accuracy in simple tests, but applying them to complex enterprise data often triggers hallucinations and incorrect logic. In fact, Stanford HAI's 2025 AI Index Report shows that some models' performance can achieve 97.8% accuracy on simpler planning tasks, but accuracy plummets when reasoning complexity increases and can reach near zero for some models.1 These errors are typically not auditable, creating a significant trust gap. "Increasingly, enterprises are embedding AI into analytics workflows to accelerate insight generation and decision-making, but speed without trust is a liability," said Guillaume Bodet, chief product officer at Actian. "Unlike error-prone text-to-SQL tools, the Actian AI Analyst is anchored in governed enterprise data and engineered for full traceability, allowing organizations to scale analytics access for business users without sacrificing accuracy or trust." The Steward Agent: Intelligently Build the Data Foundation The new Steward Agent accelerates AI deployment and replaces time-consuming manual work by intelligently building and maintaining a governed data foundation from existing warehouse schemas and business documentation. Semantic Modeling: Transforms existing business glossaries and documentation into a Semantic Knowledge Graph, enabling organizations to move from raw data to query-ready insights within days or weeks. Consistent Business Logic: Defines and maintains shared business logic to help maintain consistency across departments and prevent metric drift. Proactive Logic Maintenance: Identifies and updates changes in business definitions, allowing the semantic layer to evolve without requiring manual reconfiguration. The Actian AI Analyst: Accuracy at Scale The Actian AI Analyst delivers trusted, conversational analytics that allow business users to explore data and uncover insights using natural language. By framing every question within a governed semantic foundation, the solution enables teams to act on verified information in real time. Conversational Interface: Replaces static dashboards with a natural-language interface, allowing business users to ask follow-up questions and receive instant analysis without writing SQL. Constrained Execution Engine: Uses a rules-based approach to ground every answer in the company's specific logic and eliminates the unconstrained SQL generation that causes AI hallucinations. Full Traceability: Provides visibility into the reasoning steps, filters, and calculations used to produce every answer, allowing users to validate results and resolve discrepancies quickly. Seamless Integration and Action Actian AI Analyst integrates directly into existing business workflows to shorten the distance between a question and a strategic decision. Slack and Microsoft Teams Integration: Delivers conversational analytics directly within existing collaboration tools, allowing users to query and receive data then share results without switching applications. Threaded Conversational Context: Maintains the history of a conversation across follow-up questions, enabling insights to evolve naturally without restarting the analysis. Streamlined Operations: Empowers business users to self-serve accurate answers, reducing the operational burden on data teams. "The Actian AI Analyst turned our BI into a conversation instead of a development cycle," said Pieter Seyns, Head of Artificial Intelligence at Bekaert, a leader in material science of steel wire transformation and coating technologies. "Faster insights, fewer dashboards—the Actian AI Analyst simply works." For organizations looking to extend these capabilities enterprise-wide, Actian AI Analyst can be paired with the Actian Data Intelligence Platform to create a closed-loop system where governed data definitions flow directly into live, conversational analytics. Availability The Actian AI Analyst and Steward Agent are available now. The agents will be integrated across the Actian portfolio in the coming quarters. To learn more, read the blog and visit the Actian AI Analyst webpage. About ActianActian empowers enterprises to confidently manage and govern data at scale. Actian data management and data intelligence solutions help streamline complex data environments and accelerate the delivery of AI-ready data. Designed to be flexible, Actian solutions integrate seamlessly and perform reliably across on-premises, cloud, and hybrid environments. Learn more about Actian, the data and AI division of HCLSoftware, at actian.com. 1 Stanford University Human-Centered Artificial Intelligence (HAI), "Artificial Intelligence Index Report 2025," April 2025. Media Contact:Jennifer HarbourPR ConsultantJennifer.Harbour@actian.com Danielle LeeVice President – Global Analyst Relations & Public RelationsDanielle.Lee@actian.com
隨著生成式 AI 快速普及，企業對 AI 的關注已從技術嘗試，轉向實際導入營運流程。然而，內部資料分散、制度複雜與資安顧慮，仍使多數企業的 AI 應用停留在概念驗證階段。台灣企業數據治理與 AI 解決方案服務商 Data-DI（數據立）指出，企業 AI 能否真正落地，關鍵在於是否具備能整合既有資料、並直接銜接實際工作流程的應用層。 基於長年深耕企業數據治理與 AI 策略的實務經驗，累積服務超過百家製造、金融與公部門客戶，Data-DI 於 2025 年中推出企業級 AI Agent 平台 AltaBots.ai，協助企業快速打造可控、可用的多場景 AI 助理。實際導入數據顯示，企業平均可減少 40% 的資訊查詢時間，客服部門重複性問題處理量亦降低 50%，顯示企業 AI 應用已正式從 PoC 階段邁入規模化部署。 補上 AI 落地關鍵一環，AltaBots.ai 成為企業 AI 應用中介層 AltaBots.ai 定位為企業 AI 應用的「中介層」，可整合企業內部文件、商品資料、制度規範與歷史查詢紀錄，並彈性部署於 LINE、Slack 或企業既有系統介面中，讓 AI 直接嵌入員工日常使用的工作環境。透過多輪互動與知識理解能力，從「對話」延伸至實際執行，AltaBots.ai 協助企業建立可落地的內部知識應答與營運支援機制，而非僅停留在單點問答或展示用途。 Data-DI 解決方案顧問包威棣表示：「許多企業其實早已累積大量數位資產，但第一線人員仍花費大量時間搜尋與確認資訊。AI 的價值，在於讓正確的知識，在需要的當下被用上。」他也指出，近半年觀察到企業對 AI 的需求已明顯從「試試看」轉向「如何大規模部署」，不少客戶在 PoC 階段驗證成效後，便快速擴展至其他部門與應用場景。 AI Agent 擴展至多元企業場景，加速融入日常營運流程 目前，AltaBots.ai 已應用於多元企業場景，包括零售與大型通路的門市支援、客服與營運流程查詢，以及 HR、IT 等部門的制度與知識管理需求。透過 AI Agent 承接重複性查詢與資訊整合工作，企業得以減輕第一線與支援部門的負擔，同時提升資訊取得的一致性與效率。 為確保 AI 應用能穩定落地，Data-DI 亦搭配完整的 PoC 驗證與顧問陪跑機制，協助企業在導入初期釐清適用場景、快速啟動應用，並依實際使用情況持續優化，讓 AI 不再停留在展示或單點試用階段，而能逐步成為企業日常營運的一部分。 從顧問服務邁向平台化，Data-DI 啟動企業 AI 轉型新階段 對 Data-DI 而言，AltaBots.ai 的推出，象徵公司從顧問服務邁向產品化平台的重要里程碑，也代表企業 AI 應用正式進入可規模化部署階段。未來，Data-DI 將持續與不同產業夥伴合作，拓展 AI Agent 在營運、管理與服務流程中的應用，協助企業在風險可控的前提下，深化 AI 技術與實際流程的結合。 關於Data-DI Data-DI（數據立）是一家專注於數據治理與 AI 解決方案的企業科技服務商，致力於協助企業透過數據驅動決策、導入 AI 應用並強化組織知識資產管理。團隊擁有資訊、管理與行為科學背景，擅長將生成式 AI、語意理解與資料索引等技術應用於實際業務場景，服務涵蓋製造、金融、公共部門等多元產業。近年投入產品化策略，推出企業級 AI Agent「AltaBots.ai」，以模組化設計協助企業快速建立專屬的內部對話式應用。
全球 AI CRM 領導者 Salesforce 發布針對 2026 年的「AI 未來趨勢預測」，隨著企業 AI 應用從單純的文本生成跨入實際行動，2026 年將是「代理式企業（Agentic Enterprise）」全面落地的關鍵年。儘管 Salesforce 研究顯示 CIO 對 AI 的採用率飆升了 282%，但「數據信任」仍是企業擴展自動化的最大挑戰 Salesforce 整合內部分析與全球市場數據，歸納出 2026 年值得企業關注的三大 AI 趨勢： 趨勢一、 未來是「多代理人系統」、AI 「對結果負責」的時代 Salesforce 預測未來企業的 AI 代理人應用自主性將大幅提高，迅速轉向「精心編排的數位勞動力（Orchestrated Workforce）」模式，並且 AI 將為工作結果負起責任。所以企業需要建立多個 AI 代理人共同協作的團隊，並且 AI 代理人的角色將不只是單純接受指令，而必須為工作結果負責。如同 Slack 作為企業的代理式工作作業系統 (Agentic Work OS)，可以協助人類與多個 AI 協作，內建其中的 Slackbot 更具備了在系統中，針對不同任務自動調度不同 AI 代理人的能力。 協作取代溝通孤島：單一的 AI 代理人將成為數位孤島，難以規模化應用。代理式企業真正的成功在於建立一個「多代理人系統」，由主要的「指揮官代理人（Orchestrator Agent）」指揮其他專業代理人，執行跨部門、跨系統的協作。 成果導向：無論是提升客戶滿意度或增加銷售管道，AI 代理人將具備深度的組織意識，主動學習如何調度人員、流程與數據來達成業務成果。 事件驅動：代理人將獲得自主權，基於特定事件（如偵測到效能問題）自動觸發工作，例如自動指示開發代理人進行修復與測試，而無需人類手動提示。 人類角色的轉變：人類將轉型為高階監督者，利用 AI 可觀測性工具來設定護欄、確保道德規範，並監督整個數位團隊的績效。 趨勢二、信任即品牌：AI 體驗將決定企業競爭力 到了 2026 年，品牌的定義將不再只是商標或口號，而是取決於 AI 代理人如何代表企業的品牌，提供使用者良好的互動體驗。因此企業需要基於深度的客戶數據，打造真正理解客戶的 AI 代理人，來提供客戶高度客製化的數位互動。例如 Agentforce 平台即是深度整合 CRM 系統中的客戶數據及流暢的 AI Agent 功能，幫助 AI Agent 基於客戶資料，提供客製化服務。 AI 是終極品牌大使：那些能提供聰明、個人化且持續進化的 AI 代理人，將成為企業最強的護城河 關係智商（Relational Intelligence）：隨著消費者習慣了高度個人化的數位互動，企業將投資於更深層的關係建立。能將「建立關係」的能力移植到 AI 代理領域的企業，將獲得巨大的先行者優勢。 趨勢三、AI 將成為企業的「數位免疫系統」 企業的資安議題隨著 AI 代理人應用轉變，AI 將徹底改變網路安全的攻防戰。Salesforce 預測，資安防禦將從傳統的防火牆，轉向具備「自癒能力（Self-healing）」的企業免疫系統。因此具備自動偵測與自動修復的 AI 代理人，將成為企業資安的重要防線。 像是 Salesforce 透過 可信賴的 AI 基礎 (Trusted AI Foundation) 為企業的 AI 代理人建立穩固的資安防線，讓 AI 代理人能在安全的邊界（Trust Boundary）內運作，既能深度讀取企業數據提供精準回覆，也能確保 AI 代理人的每一項操作都符合資料治理規範，真正實現安全、可量產的 AI 生產力。 主動式獵捕：AI 將協助資安團隊從被動反應轉為主動預測，自動化處理重複性任務並即時偵測異常。 自動修復：自主安全代理人將監控業務流程的「健康狀況」，一旦偵測到威脅偏離常態，能在毫秒內自動隔離受感染的工作流並修補漏洞，無需等待人類批准即可阻斷攻擊。 Salesforce 致力於透過可信賴的統一平台，協助全球企業組織整合人類、AI 代理人與數據，加速轉型為「代理式企業」。秉持深耕在地的承諾，Salesforce 台灣亦迎來嶄新里程碑：即日起正式更名為「台灣賽富時有限公司（Salesforce Taiwan Ltd.）」，位於台北 101 大樓的新辦公室亦於今年正式啟用。展望未來，Salesforce 將以新據點為核心，結合全球前瞻的 AI 趨勢洞察與強大的在地團隊，全方位協助台灣企業在 AI 新時代中搶佔先機。
A12 藝術空間
Slack
請先登入後才能發佈新聞。
還不是會員嗎？立即 加入台灣產經新聞網會員 ，使用免費新聞發佈服務。 （服務項目） （投稿規範）
