麥肯錫 2026 全球 AI 調查：大企業代理規模化達 40%，但 EBIT 貢獻仍停在 37%
The state of AI in 2026: On the road to ROI
摘要分析
麥肯錫 2026 年 AI 現況調查（1,719 名受訪者、97 國）顯示採用深化但財務成效持平。年營收逾 10 億美元的大型組織有 40% 已規模化 AI 代理（去年 27%），較小組織維持 22%；32% 因能用程式代理自行開發而放棄採購軟體。約 20% 組織表示 AI 營運成本已限制使用，但 60% 仍計畫加碼投資。八成受訪者稱 AI 提升個人生產力，但僅 37% 將 EBIT 貢獻歸功於 AI，與去年相同，高績效組織仍約 6%。39% 預期明年 AI 將減少人力，但去年預期裁員的 32% 中實際發生的僅 14%。
- 大企業代理規模化從 27% 升至 40%，小企業停在 22%
- 32% 企業因能自建而放棄採購軟體
- 20% 組織受 AI 營運成本限制，60% 仍要加碼
- 80% 個人生產力提升，但 EBIT 貢獻僅 37% 未變
- 預期裁員 39%，實際發生僅 14%
McKinsey's 2026 State of AI survey (1,719 respondents in 97 nations, fielded May 4–June 8, 2026) finds adoption deepening but financial impact flat. Forty percent of respondents at organizations with over $1 billion in revenue report scaling AI agents, up from 27%, while smaller organizations stayed at 22%; about two in ten overall are scaling agents and a similar share software coding agents (31% at large enterprises). Nearly a third (32%) decided against buying at least one software product because they could build it in-house with coding agents. AI-related operating costs now constrain usage for about 20% of organizations, yet 60% plan to increase investment. Eighty percent say AI improved their personal productivity and half say it improves decisions, but only 37% attribute any EBIT impact to AI (unchanged) and high performers remain about 6%. Thirty-nine percent expect AI-driven headcount declines next year, though only 14% saw them last year versus 32% who had predicted them.
- Large firms scaling agents rose from 27% to 40%; smaller firms flat at 22%
- 32% skipped buying software because coding agents built it in-house
- 20% constrained by AI operating costs; 60% plan to invest more
- 80% report personal productivity gains but EBIT impact stuck at 37%
- 39% expect AI-driven headcount cuts; only 14% actually saw them
關鍵字：麥肯錫AI 代理企業採用投資報酬程式代理
來源原文：https://www.mckinsey.com/capabilities/quantumblack/our-insights/the-state-of-ai
本頁為報導內容的摘要與分析，版權屬原媒體所有；請以原文為準。