I’ve always been somewhat skeptical of overly using the word “autonomous” when applied too liberally. With AI’s growing capabilities and more enterprises deploying AI agents, the question is whether the autonomous enterprise is here and now.

And what should an autonomous enterprise mean?

Does it mean workflows will become largely human-less, with AI agents not only taking on tasks but also conversing through integrated MCP services to orchestrate more complex workflows?

I remain skeptical whether leaders will fully trust fully automated agentic experiences. I’ve even advised CIOs to avoid using the word ‘automation’ when describing business processes, RPAs (remote process automations), and data pipelines. Automation sets a high bar for what systems and AI will do for enterprises and may strike fear into employees if they believe it will lead to layoffs and their own job loss.

But autonomous is the future – and paving the way for how the autonomous enterprise will augment people’s creativity, team productivity, and growth should be top of mind for every CIO.  

The brain behind the autonomous enterprise

At SAP Sapphire 2026, CEO Christian Klein introduced their updated vision for SAP and its customers – the autonomous enterprise – proclaiming the ERP is the brain providing the context behind agentic experiences. The context leads to an explosion of AI agents that SAP customers can use out of the box, extend with their own business rules, develop themselves, or integrate via MCP servers.

Let me unpack this for you – here’s what CIOs need to know.

When autonomous drives employee adoption

SAP’s leadership team introduced 40 Joule agents at SAP Sapphire 2025 one year ago. Today, SAP has 200+ out-of-the-box Joule agents – and some enterprises have deployed many more through SAP and other platforms. One chief digital officer I met at SAP Sapphire 2026 told me his retail company has already deployed over 1,000 AI agents.

To accelerate adoption, SAP introduced a taxonomy to help leaders and employees modernize to agile workflows. Joule Agents are grouped into 50+ Joule Assistants, which are organized into five domains: finance, spend, supply chain management, human capital management, and customer experience (CX).

For example, the CX domain includes the following assistants: deal qualification, content, campaign, sales, deal closing, shopping, merchandising, order management, self-service, and case management.

Why this matters: Names matter – calling them role-based assistants will help grow adoption.

The word ‘assistant’ implies that the underlying AI agents intend to partner with and augment employee capabilities. They are intentionally role-based, so employees can easily find which agents to try and learn their capabilities. Assistant plays off the familiar virtual assistant term, but because they are role-based, it sets a higher expectation on value and outcomes delivered. Some metrics SAP shared about the CX virtual assistants are a 22% increase in expansion revenue rate and a 15% increase in conversion from consideration to cart.

Leading with governance-first

Last year’s challenge was selecting AI experiments that matter. This year’s opportunity is deploying AI agents, growing adoption, and delivering outcomes from the investment.

But as more CIOs deploy hundreds to thousands of AI agents, governing and managing them will be a growing responsibility.

SAP introduced several agent governance capabilities during Sapphire. SAP AI Agent Hub is the catalog of deployed SAP and third-party agents. In addition to helping employees discover agents, the hub provides controls for admins to manage them. Their agent monitoring tracks task success, latency, output correctness, and rule compliance.

Why this matters: Governance and management must outpace value and innovation.

CIOs are under pressure to deliver ROI from AI investments. The companies showing early success in deploying AI agents are quickly realizing the importance of continuous contextual monitoring of the AI’s outputs. Centralizing an AI hub for access, activity, and management is a fundamental tool for growing smart, safe adoption.

Building and extending AI agents

SAP introduced Joule Studio 2.0 – the crown jewel of their Sapphire 2026 announcements. It’s free to develop (for now) so that SAP customers can learn how to build, test, and deploy AI agents without worrying about token costs.

Among Joule Studio’s new capabilities:

• “Enterprise vibe coding,” starting with intent/requirements and leading to testing/deployment.
• Review the product requirements document (PRD), visualize the solution, and develop testing evals.
• Built on an open model, an open code-generating tool, and an open cloud deployment providing architecture, compliance, and developer experience flexibility.
• Leverage Industry AI Solutions in 20+ industries.

Why this matters:  AI agent options flexibility

Teams can explore several options: use out-of-the-box AI agents, leverage their innovative capabilities to extend SAP’s AI agents with proprietary rules and knowledge, or develop proprietary AI agents. SAP follows a spec-driven development approach with more advanced testing and specification capabilities than I’ve seen on most other platforms.

Why CIOs should review SAP’s AI capabilities

SAP is showing customers why they should trust developing AI capabilities with their platforms and partners. Their commitment to developing the autonomous enterprise with “assistants” should improve adoption, while their open strategy provides flexibility in model, development, and deployment.