2025年人工智能的现状：代理、创新与转型

Agents, innovation, and transformation November 2025 Almost all survey respondentssay their organizations are usingAI, and many have begun to useAI agents. But most are still inthe early stages of scaling AI andcapturing enterprise-level value. Key findings 1.Most organizations are still in the experimentation or piloting phase:Nearly two-thirdsof respondents say their organizations have not yet begun scaling AI across the enterprise. 2.High curiosity in AI agents:Sixty-two percent of survey respondents say theirorganizations are at least experimenting with AI agents. 3.Positive leading indicators on impact of AI:Respondents report use-case level cost andrevenue benefits, and 64 percent say that AI is enabling their innovation. However, just 4.High performers use AI to drive growth, innovation, and cost:Eighty percent ofrespondents say their companies set efficiency as an objective of their AI initiatives,but the companies seeing the most value from AI often set growth or innovation as 5.Redesigning workflows is a key success factor:Half of those AI high performers intendto use AI to transform their businesses, and most are redesigning workflows. 6.Differing perspectives on employment impact:Respondents vary in their expectationsof AI’s impact on the overall workforce size of their organizations in the coming year: Three years sincethe introduction of gen AI toolstriggered a new era of artificial intelligence, nearly nineout of ten survey respondents say their organizations into their workflows and processes to realize material enterprise-level benefits. The latest McKinsey Global Survey on the state of AIreveals a landscape defined by both wider use—including growingproliferation of agentic AI—and stubborn growing pains, with the AI use continues to broaden butremains primarily in pilot phases Our latest survey shows a larger share of respondents reporting AI use by their organizations,though most have yet to scale the technologies. The share of respondents saying theirorganizations are using AI in at least one business function has increased since our researchlast year: 88 percent report regular AI use in at least one business function, compared with78 percent a year ago. But at the enterprise level, the majority are still in the experimenting or Exhibit 1Web <2025><StateofAI2025>Exhibit <1> of <20>Reported use of AI in at least one business function continues to increase. Use of AI by respondents’ organizations,% of respondents Organizations that use AI in at least 1 business function¹ McKinsey & Company Many organizations are already experimenting with AI agents Organizations are also beginning to explore opportunities with AI agents—systems based onfoundation models capable of acting in the real world, planning and executing multiple steps in aworkflow. Twenty-three percent of respondents report their organizations are scaling an agenticAI system somewhere in their enterprises (that is, expanding the deployment and adoption ofthe technology within a least one business function), and an additional 39 percent say they have Web <2025><StateofAI2025> Phase of AI agent use at respondents’ organizations, by business function,¹% of respondents (n = 1,933) McKinsey & Company Looking at individual business functions, agent use is most commonly reported in IT andknowledge management, where agentic use cases such as service-desk management in IT anddeep research in knowledge management have quickly developed. By industry, the use of AI Exhibit 3Web <2025><StateofAI2025>Exhibit <3> of <20> Use of AI agents is most often reported by respondents working intechnology, media and telecommunications, and healthcare. AI agent use that has reached the scaling phase,1by industry and business function,% of respondents McKinsey & Company McKinsey commentary Michael Chui Senior fellow AI agents have been the subject of intense buzz and excitement. Already, about a quarter ofour survey respondents report that they have started scaling at least one agentic AI system,but usually only in one or two business functions. Looking across the entire enterpriselandscape, the use of agents is not yet widespread. This gap highlights the contrast between the great potential that manifests in a “hype cycle” and the current reality on the ground: Forthose companies that respondents say have started to use agents in any particular business Twenty-three percent of respondentsreport their organizations are scalingan agentic AI system somewhere For most organizations, AI use remains in pilot phases The use of AI overall is broadening within organizations. Respondents increasingly report thattheir organizations are using AI in more business functions (Exhibit 4). More than two-thirds ofrespondents now say their organizations are using AI in more than one function, and half reportusing AI in three or more functions (for a breakdown by industry, see sidebar, “Reported AI use Web <2025><StateofAI2025> Business fu