从辅助到执行：企业如何让AI投入工作

各类组织正在扩大人工智能的应用范围，并提升对其任务要求的复杂度。企业级AI正从辅助角色转向执行角色，但并非所有企业都在以相同速度推进这一转变。前沿企业——即每月AI使用量排名前10%的企业——现在每位活跃用户生成的输出令牌数已达到普通企业的8.3倍。这一指标是使用深度的代理衡量标准，而差距的扩大伴随着更多将智能体连接至公司上下文、工具和可重复工作流程的能力被采纳。

今天，我们发布两份互补的研究来审视这一转变。Enterprise Signals 以实用视角切入OpenAI企业客户群中的智能体AI应用，包括前沿企业的差异化做法以及智能体工作正在扩展的领域。配套工作论文《组织如何使用AI：来自ChatGPT的证据⁠(在新窗口中打开)》(https://cdn.openai.com/pdf/how-organizations-use-chatgpt.pdf) 则考察了AI采纳如何在不同公司、角色和资历层级间增长。

综合来看，这两项研究指向一个务实的企业议程：将智能体连接到完成有价值工作所需的上下文和工具；建立明确的权限、审查和治理机制；并帮助员工将有效的个人工作流程转化为共享的工作方式。

这些报告揭示了关于组织如何应用AI的五大关键洞察：

• 企业使用正变得更加智能体化。 截至6月，在企业客户中，Codex生成的输出令牌占Codex和ChatGPT合计输出令牌的64%，这表明智能体正在推动向更实质性、可委派工作的转变。
• 前沿差距正在扩大。 前沿企业——即每月AI使用量排名前10%的企业——现在每位活跃用户生成的输出令牌数已达到普通企业的8.3倍，而1月份这一数字为2.6倍。
• 前沿企业更频繁地使用高级功能。 每周，前沿企业中有21%的活跃用户使用插件，而普通企业仅为9%。在OpenAI内部，95%的员工每周使用插件，凸显了更深层次采纳的潜力。
• 智能体正在知识工作中广泛扩展。 自2月以来，企业级Codex的每周活跃用户在法务领域增长了108倍，销售领域增长41倍，招聘领域增长41倍，市场营销领域增长26倍，而工程领域仅增长5倍。
• 早期职业员工更多使用AI。 使用率在早期职业员工中最高，在更资深的员工中则有所下降，这表明他们在使用AI方面可能具有比较优势。

企业正从“询问”转向“执行”

企业级AI正从回答问题转向执行工作。助手帮助人们思考工作；智能体则帮助他们完成工作。ChatGPT Work和Codex等产品可以使用工具、创建文件并产出供审查的工作成果。例如，员工不再需要询问AI如何准备演示文稿，而是可以要求智能体跨来源收集相关信息并直接起草演示文稿本身。

这一转变在企业使用中清晰可见。截至6月，在企业客户中，Codex生成的输出令牌占Codex和ChatGPT合计输出令牌的64%。智能体工作流程通常会产生更多输出，因为它们执行更长的多步骤任务，因此这一数字既反映了Codex的使用频率，也反映了这些任务产生的输出量。

随着企业拥抱智能体AI，前沿差距扩大

每个月，我们根据每位活跃用户的输出令牌数对企业客户进行排名。前沿企业是当月排名前10%的企业，而普通企业则位于第45至第55百分位之间。截至6月，前沿企业每位活跃用户生成的输出令牌数是普通企业的8.3倍，与1月份2.6倍的差距相比增长了三倍。前沿差距出现在各个行业和不同规模的公司中，表明密集型AI使用并不局限于科技公司。Enterprise Signals 考察了这一差距在不同行业和职能间的差异。

在《组织如何使用AI：来自ChatGPT的证据⁠(在新窗口中打开)》(https://cdn.openai.com/pdf/how-organizations-use-chatgpt.pdf) 所研究的美国上市公司中，企业级采纳者相比非采纳者拥有更强的财务指标。它们持有更多资产，雇佣更多员工，并拥有更高的研发投资水平。综合来看，这些报告表明，仅凭访问权限可能不足以扩展AI应用。在持续员工学习、共享工作流程、数据基础设施和治理方面的配套投资，可以支持更广泛和更深入的采纳。

插件和技能等高级功能在前沿企业中更为普遍

AI智能体需要访问正确的上下文和工具才能有效运作。插件⁠(在新窗口中打开) 将帮助智能体完成特定工作流程的能力捆绑在一起。它们可以将提供可复用指令的_技能_与连接公司数据、工具和操作的_应用_相结合。例如，销售插件可以将团队的行动手册与其CRM访问权限结合，使智能体能够利用当前客户信息和过往提案来准备定制化回复供审查。

前沿企业在高级功能方面具有明显领先优势。在每周活跃用户中，前沿企业有21%使用插件，19%使用技能，而普通企业仅为9%和3%。然而，前沿企业的采纳率仅代表了可能性的冰山一角。OpenAI的内部使用情况凸显了这些功能更深层使用的潜力，每周插件使用率达到了活跃用户的95%。我们关于智能体如何改变工作的研究更深入地展示了OpenAI员工如何使用高级功能。

智能体正在知识工作中广泛扩展

软件工程是智能体采纳的早期中心，但Codex的使用现在正在各知识工作职能中快速增长。自2月以来，企业级Codex的每周活跃用户在法务领域增长了108倍，销售领域增长41倍，招聘领域增长41倍，市场营销领域增长26倍，而工程领域仅增长5倍。

在维珍大西洋航空，这一转变在整个业务中清晰可见。工程团队使用Codex在30分钟内重构遗留代码，而过去需要两周。他们的产品团队使用ChatGPT Work在数小时内完成原本需要数周的竞争研究，从而塑造了该航空公司的五年数字战略。

Enterprise Signals 基于超过1000万条消息的样本，探讨了这些能力如何在各行业和职能间扩展。

早期职业员工更多使用AI

使用率在早期职业员工中最高，在更资深的员工中则有所下降。这一模式表明，早期职业员工可能在利用AI方面具有比较优势。

多项调查显示，领导者和高管层使用人工智能的比例更高。然而，来自数百万次对话的管理数据却得出了相反的结论。在采用AI六个月后，早期职业员工每周发送的消息比高管多出13条。

对于领导者而言，这一结果指向了一个实际机会：识别出具有最强AI使用习惯的员工，并将其工作流程公开化，从而帮助有效实践在组织各层级中推广。

领导者如何缩小前沿差距

各公司或许都能使用相同的模型，但前沿企业正在更快速、更深入地将这些模型应用于整个组织。

领导者的机遇在于，通过将代理式工作流程扩展到工程部门之外，并将成功的个人用例转化为组织内可复制的实践，从而缩小前沿差距。通过将代理连接到公司上下文和工具，并明确权限、治理和人工审核机制，企业可以从辅助执行迈向自主执行。

阅读 Enterprise Signals 以深入了解前沿企业如何跨行业、跨职能应用AI。OpenAI企业客户还可以 申请 定制基准测试，以了解其组织与前沿企业的对比情况。

Organizations are expanding both where they use AI and what they ask it to do. Enterprise AI is moving from assistance to execution, yet not all firms are making that transition at the same pace. Frontier firms—those in the top 10% of AI usage each month—now generate 8.3× as many output tokens per active user as typical firms. The measure is a proxy for depth of use, and the widening gap appears alongside greater adoption of capabilities that connect agents to company context, tools, and repeatable workflows.

Today we are publishing two complementary studies that examine this shift. Enterprise Signals leads with a practical view of agentic AI across OpenAI’s enterprise customer base, including what frontier firms are doing differently and where agentic work is spreading. The companion working paper, How Organizations Use AI: Evidence from ChatGPT⁠(opens in a new window), examines how adoption grows across companies, roles, and levels of seniority.

Together, the studies point to a practical enterprise agenda: connect agents to the context and tools needed to complete valuable work; establish clear permissions, review, and governance; and help employees turn effective individual workflows into shared ways of working.

These reports reveal five key insights on how organizations are putting AI to work:

• **Enterprise use is becoming more agentic.**As of June, Codex generated 64% of combined Codex and ChatGPT output tokens among enterprise customers, suggesting that agents are enabling a shift toward more substantive, delegated work.
• **The frontier gap is widening.**Frontier firms—those in the top 10% of AI usage each month—now generate 8.3× as many output tokens per active user as typical firms, up from 2.6× in January.
• **Frontier firms use advanced capabilities more often.**Each week, 21% of active users at frontier firms use Plugins, compared with 9% at typical firms. At OpenAI, 95% of employees use Plugins weekly, highlighting the potential for deeper adoption.
• **Agents are spreading across knowledge work.**Since February, weekly active enterprise Codex users grew 108× in legal, 41× in sales, 41× in recruiting, and 26× in marketing, compared with 5× in engineering.
• **Early-career employees use AI more.**Usage is highest among early-career workers and falls among more senior employees, suggesting a potential comparative advantage in using AI.

Enterprises are shifting from asking to doing

Enterprise AI is moving from answering questions to carrying out work. Assistants help people think through work; agents help them complete it. Products like ChatGPT Work and Codex can use tools, create files, and produce work for review. Instead of asking AI how to prepare a presentation, for example, a worker can ask an agent to gather relevant information across sources and draft the presentation itself.

This shift is visible in enterprise usage. As of June, Codex generated 64% of combined Codex and ChatGPT output tokens among enterprise customers. Agentic workflows typically generate more output because they carry out longer, multi-step tasks, so the figure reflects both how often Codex is used and how much output those tasks produce.

The frontier gap widens as firms embrace agentic AI

Each month, we rank enterprise customers by output tokens per active user. Frontier firms are those in the top 10% that month, while typical firms fall between the 45th and 55th percentiles. As of June, frontier firms generated 8.3× as many output tokens per active user as typical firms, a threefold increase compared to the 2.6× gap in January. The frontier gap appears across industries and company sizes, showing that intensive AI use is not limited to technology companies. Enterprise Signals examines how the gap varies across industries and functions.

Among the U.S. public companies studied in How Organizations Use AI: Evidence from ChatGPT⁠(opens in a new window), enterprise adopters had stronger financial measures compared to non-adopters. They held more assets, employed more workers, and had higher levels of R&D investment. Read together, the reports suggest that access alone may not be enough to scale AI. Complementary investments in continuous employee learning, shared workflows, data infrastructure, and governance can support broader and deeper adoption.

Advanced capabilities like Plugins and skills are more common at frontier firms

AI agents need access to the right context and tools to be effective. Plugins⁠(opens in a new window) bundle capabilities that help agents complete specific workflows. They can combine skills that provide reusable instructions with apps that connect to company data, tools, and actions. For example, a sales Plugin can combine a team’s playbook with access to its CRM, allowing an agent to use current customer information and past proposals to prepare a tailored response for review.

Frontier firms have a clear lead in advanced capabilities. Among weekly active users, 21% at frontier firms use Plugins and 19% use skills, compared with just 9% and 3% at typical firms. However, frontier firm adoption represents only a fraction of what is possible. OpenAI’s internal usage highlights the potential for deeper usage of these capabilities, with weekly Plugin usage at 95% of active users. Our research on how agents are transforming work provides a closer look at how employees at OpenAI use advanced capabilities.

Agents are spreading across knowledge work

Software engineering was an early center of agentic adoption, but Codex use is now growing quickly across knowledge-work functions. Since February, weekly active enterprise Codex users grew 108× in legal, 41× in sales, 41× in recruiting, and 26× in marketing, compared with 5× in engineering.

At Virgin Atlantic, that shift is visible across the business. Engineering teams use Codex to refactor legacy code in 30 minutes instead of two weeks. Their product teams use ChatGPT Work to complete weeks of competitive research in hours, shaping the airline’s five-year digital strategy.

Enterprise Signals explores how these capabilities are spreading across industries and functions, drawing on a sample of more than 10 million messages.

Early-career employees use AI more

Many surveys have reported higher levels of AI use among leaders and executives. However, administrative data from millions of conversations finds the opposite. Six months after adoption, early-career employees sent 13 more messages per week than executives.

For leaders, the result points to a practical opportunity to identify employees with the strongest AI habits and make their workflows visible, helping effective practices spread across all levels.

What leaders can do to narrow the frontier gap

Companies may have access to the same models, but frontier firms are putting them to work faster and more deeply across their organizations.

The opportunity for leaders is to close the frontier gap by extending agentic workflows beyond engineering and turning successful individual use cases into repeatable practices across the organization. By connecting agents to company context and tools, with clear permissions, governance and human review, firms can move from assistance to execution.

Read Enterprise Signals for deeper insights into how frontier firms are putting AI to work across industries and functions. OpenAI Enterprise customers can also request a customized benchmark to see how their organization compares against frontier firms.