How

enterprises

are

building

AI

agents

in

2026

AI agents have moved from experimental to an essential part of an organization’s tech stack. But how are enterprises actually using them?AI代理已从实验阶段转变为组织技术栈中不可或缺的一部分。但企业实际上是如何使用它们的呢？

In partnership with research firm Material, we surveyed over 500 technical leaders across industries and company sizes to understand how organizations are deploying agents today and where they see opportunity ahead.我们与研究公司Material合作，调查了来自不同行业和公司规模的500多名技术领导者，以了解组织目前如何部署代理以及他们未来看到的机会。

The findings show a clear pattern: organizations are shifting from simple task automation to complex, multi-step workflows that span teams and business functions.调查结果显示出清晰的模式：组织正在从简单的任务自动化转向跨越团队和业务职能的复杂多步骤工作流。

What the data shows数据表明

More than half of organizations (57%) now deploy agents for multi-stage workflows, with 16% running cross-functional processes across multiple teams. In 2026, 81% plan to tackle more complex use cases, including 39% developing agents for multi-step processes and 29% deploying them for cross-functional projects.超过一半的组织（57%）现在部署代理用于多阶段工作流，其中16%运行跨职能流程。到2026年，81%计划应对更复杂的用例，包括39%开发用于多步骤流程的代理，29%部署用于跨职能项目。

Coding leads adoption. Nearly 90% of organizations use AI to assist with development, and 86% deploy agents for production code. Organizations report time savings across the entire development lifecycle: planning and ideation (58%), code generation (59%), documentation (59%), and code review and testing (59%).编码引领采用。近90%的组织使用AI辅助开发，86%部署代理用于生产代码。组织报告在整个开发生命周期中节省时间：规划和构思（58%）、代码生成（59%）、文档编写（59%）以及代码审查和测试（59%）。

But the impact extends beyond engineering. Data analysis and report generation (60%) and internal process automation (48%) rank among the highest-impact use cases. Looking ahead, 56% plan to implement agents for research and reporting over the next year.但影响不仅限于工程领域。数据分析和报告生成（60%）以及内部流程自动化（48%）位列影响最大的用例。展望未来，56%计划在未来一年内实施代理用于研究和报告。

Perhaps most notably, 80% of organizations report their AI agent investments are already delivering measurable economic returns. 最值得注意的是，80%的组织报告其AI代理投资已经带来可衡量的经济回报。

What this looks like in practice实际应用案例

The organizations seeing results are treating agents as a core part of their infrastrucuture, not experiments.取得成果的组织将代理视为其基础设施的核心部分，而非实验。

Thomson Reuters uses Claude to power CoCounsel, their AI legal platform. Lawyers who once spent hours manually searching through documents can now access 150 years of case law and 3,000 domain experts in minutes. 汤森路透使用Claude为其AI法律平台CoCounsel提供支持。曾经花费数小时手动搜索文档的律师现在可以在几分钟内访问150年的判例法和3000名领域专家。

Cybersecurity company eSentire compressed expert threat analysis from 5 hours to 7 minutes, with AI-driven analysis aligning with their senior security experts 95% of the time. In healthcare, Doctolib rolled out Claude Code across their entire engineering team, replacing legacy testing infrastructure in hours instead of weeks and shipping features 40% faster.网络安全公司eSentire将专家威胁分析从5小时压缩到7分钟，AI驱动的分析与高级安全专家的一致性达到95%。在医疗领域，Doctolib在其整个工程团队中推广Claude Code，在数小时内取代了传统测试基础设施，而不是数周，功能交付速度提高了40%。

The retail sector is seeing similar gains. L'Oréal achieved 99.9% accuracy on conversational analytics, enabling 44,000 monthly users to query data directly instead of waiting for custom dashboards.零售行业也看到了类似的收益。欧莱雅在对话分析上实现了99.9%的准确率，使每月4.4万用户能够直接查询数据，而无需等待定制仪表板。

The path forward前进之路

The question for leaders in 2026 isn't whether to adopt AI agents but how to scale them strategically. The data points to three primary challenges: integration with existing systems (46%), data access and quality (42%), and change management needs (39%).2026年领导者面临的问题不是是否采用AI代理，而是如何战略性地扩展它们。数据指出了三个主要挑战：与现有系统的集成（46%）、数据访问和质量（42%）以及变更管理需求（39%）。

Nine in 10 leaders report that agents are shifting how their teams work, with employees spending more time on strategic activities, relationship building, and skill development rather than routine execution.十分之九的领导者报告称，代理正在改变团队的工作方式，员工将更多时间用于战略活动、关系建设和技能发展，而非常规执行。

This transition requires purpose-built infrastructure: models optimized for coding and enterprise workflows, frameworks like the Agent SDK, and tools like Claude Code that help teams move from prototype to production faster. 这种转变需要专门构建的基础设施：针对编码和企业工作流优化的模型、像Agent SDK这样的框架，以及像Claude Code这样的工具，帮助团队更快地从原型过渡到生产。

We're also finding that while coding has been the proving ground for AI agents,  it's just the beginning. As agents expand into research, customer service, financial planning, and supply chain operations, the organizations that build expertise now will capture disproportionate value as the technology matures.我们还发现，虽然编码一直是AI代理的试验场，但这仅仅是个开始。随着代理扩展到研究、客户服务、财务规划和供应链运营，现在建立专业知识的组织将在技术成熟时获得不成比例的价值。

Read the full 2026 State of AI Agents Report, here.在此处阅读完整的《2026年AI代理现状报告》。