2025全球AI就绪指数报告：企业中AI智能体的规模化应用

ForewordThe rapid evolution of Artificial Intelligence (AI) presents an unparalleled opportunity to reshape ourworld. It will drive innovation, enhance productivity, and fundamentally transform how businessesoperate and governments serve their citizens. At Salesforce, we believe that the next wave of thisrevolution, agentic AI, holds the potential to unlock even greater opportunities for growth, impact, andefficiencies and ultimately create profound societal value by fostering human-AI collaboration in newways and enabling AI systems to more autonomously plan, reason, and act to advance organizationalpriorities and complete complex tasks on behalf of humanity.This potential transformation hinges not only on technological advancement but also on establishing afoundation of trust and responsible governance. This Global AI Agents Readiness Index is a contributionto this important discussion. It offers a comprehensive framework for understanding how preparedcountries are to harness agentic AI, highlighting the diverse approaches to regulation, diffusion,investment, and talent development across sixteen key global markets.What this report makes clear is that countries are eager to advance in their transformation journeys. Whilepolicymakers have often focused on risk identification and management, there is another imperative:actively enabling the strategic adoption of trusted agentic AI. Future-forward countries that prioritizeinvestment in innovation, infrastructure, talent, upskilling and reskilling, and ethical frameworks will bebest positioned to compete, lead in this new era of digital transformation, and navigate thoughtfully thecoming future of human-AI collaboration.Salesforce is deeply committed to fostering trusted agentic AI. Our approach to governance, guardrails,and guidance – including trust layers and mechanisms for visibility, observability and transparency,embedded within solutions like Agentforce – ensures that every interaction is governed by robust dataprivacy and security controls, reinforcing our number one value: trust. We believe that by building trustinto the core of these technologies, we can empower organizations to deploy AI agents confidently andsecurely, unleashing their full enterprise value.This Index serves as a valuable resource for governments, businesses, and civil society, providing insightsinto the strengths and challenges in scaling agentic AI. It underscores the importance of continuedinvestment, cross-sector collaboration, and strategic focus to accelerate adoption and ensure thatthe benefits of agentic AI are made accessible, scalable, and inclusive for all. Together, we can seizethis opportunity to create a future where humans and AI agents together streamline public services,strengthen economic competitiveness, and contribute to a more effective, efficient, and responsiveglobal landscape.Sabastian V. NilesPresident and Chief Legal OfficerSalesforce, Inc. 2 Contents4Executive Summary9Introduction13Methodology15Key Findings24Recommendations29Looking Ahead:Seizing theAgentic AI Opportunity30Appendix I: Detailed Findings34Appendix II: Methodology Executive SummaryThe adoption of artificial intelligence (AI), and in particular agentic AI, presents asignificant opportunity for both the public and private sectors. These technologies cantransform how services are delivered, how businesses operate, and how governmentsengage with citizens — unlocking new efficiencies, driving economic growth, andenhancing competitiveness.In recent years, the AI revolution has developed through three waves: predictive AI,generative AI, and now agentic AI. Agentic AI is the technology that powers AI agentsso they can act autonomously without human oversight. What sets autonomous agentsapart from their predecessors is that they can autonomously plan, reason, and act tocomplete complex tasks. The adoption of AI agents, working hand-in-hand with humanemployees,is expected to increase by 327%over the next two years, leading to a30% productivity gain.As agentic AI and its applications mature, countries that are prepared toharness their potential will be best positioned to lead in the next wave of digitaltransformation.To this end, governments worldwide are exploring diverse governanceapproaches, ranging from comprehensive regulatory frameworks to more flexible,industry-led approaches.•In the European Union (EU), member states like France, Germany, and Italy willbe applying theEU AI Act, a comprehensive, risk-based regulatory frameworkthat categorizes AI systems based on their potential risks — imposing strictercompliance requirements on those deemed high-risk.•South Korea has adopted a regulatory and governance structure in its recentBasic Law on the Development of AI, combining requirements for certainhigh-risk systems with measures to promote development and adoption of AI.•Japan, with its recently adoptedBill on the Promotion of Research,Development and Utilization of Artificial Intelligence‑Relat