《中国企业智能化成熟度报告(2025)——企业智能化迈向AI原生新时代》
Search documents
联想发布《中国企业智能化成熟度报告》 企业智能体呈现规模化落地趋势
Zheng Quan Ri Bao Wang· 2026-02-12 07:14
Group 1 - The core viewpoint of the report is that Chinese enterprises are transitioning towards an AI-native era, with AI concepts increasingly integrated into corporate strategies, leading to a trend of large-scale implementation of intelligent systems [1][2] - By 2025, the proportion of enterprises at advanced stages of intelligent transformation (levels four and five) is expected to rise significantly to 39%, with AI-native enterprises making up 9% of this figure, compared to 16%, 22%, and 22% in 2022, 2023, and 2024 respectively [1] - The overall average maturity score across industries is projected to reach 3.19, a notable increase from 2.77 in 2024, with the financial sector maintaining the highest maturity score [1][2] Group 2 - The financial industry continues to lead with a maturity score of 3.43, with 49% of enterprises in levels four and five by 2025, up from 30% in 2024 and 20% in 2022 [2] - The healthcare sector has the highest proportion of leading enterprises at 50%, transitioning from pilot phases to advanced stages of "intelligent operations + AI-native" [2] - The manufacturing sector remains at a low maturity level, slightly below the industry average, indicating significant room for improvement due to disparities in complexity and digital foundations among its sub-sectors [2] Group 3 - The report emphasizes a transformation framework focused on "value-driven, systematic advancement," highlighting three main values: operational value, strategic value, and industry and social value [3] - Enterprises have made significant progress in their intelligent transformation, establishing clear digital visions and unified transformation blueprints, supported by dedicated investments and change management mechanisms [3] - AI-native enterprises are shifting from "innovation experiments" to mainstream models, aiming to reconstruct the entire value chain from strategic planning to execution with an AI-first approach [3]