On July 2, 2026, SAP sent an internal memo to its entire workforce. Bloomberg saw the document. The contents were blunt: all hiring outside of AI roles is paused. Internal travel unrelated to AI development or customer work is suspended. Supplier spending is under review. The only category getting more money is artificial intelligence.

SAP has more than 100,000 employees. Its software processes 77% of the world’s transaction revenue. It runs the back office of most Fortune 500 companies. Its restructuring has implications for those customers too.

Christian Klein told The New York Times he expects the workforce to remain at a similar size while its composition changes substantially. His expectation points to a broad rewriting of job descriptions.

The Restructuring Behind the Restructuring

Between 2023 and 2024, SAP cut roughly 12,000 positions in a restructuring that cost over three billion euros. The current memo redirects hiring and spending toward AI.

The difference matters. Layoffs are subtraction. What SAP is doing now is substitution. Every role that opens, every dollar that gets spent, every trip that gets approved runs through a single filter: does this advance agentic AI?

SAP is rebuilding its operating structure around AI alongside its product roadmap. The agentic software it builds next will be shaped by that workforce redirection.

The memo gives a hundred thousand people a direct instruction to align their work with the company’s AI direction.

What SAP’s Direction Means for Customers

If your company runs SAP for ERP, HR, supply chain, finance, or procurement, its direction will shape the tools your team relies on.

SAP is building agentic software into the tools teams already use every day. Joule, SAP’s AI assistant, is already embedded in most of its cloud products. The direction is toward agents that can take action inside those workflows.

That means the organizations running SAP will get pushed into agentic workflows whether they planned for it or not. The question is whether you are ready when that push arrives.

The Kyndryl People Readiness Report from July 2026 found that AI deployment hit 57% of enterprises, up from 35% a year ago. But only 11% of those enterprises are meeting their AI objectives. That leaves a substantial gap between deployment and reported results.

The vendor that runs your back office is telling its own people to restructure around AI. Its customers should examine what preparation their own teams need.

The Gap Just Changed Shape

AI readiness depends on how organizations redesign operations around the technology. Look at whether decisions, responsibilities, and handoffs support agent work inside processes originally designed for humans working alone.

SAP’s memo reveals another planning constraint: your vendor sets its own timeline. Prepare your workflows against that roadmap and identify where your team needs time to adapt.

This is how compounding works. The first wave was companies that moved early getting ahead. The second wave is the infrastructure shifting underneath everyone else, pulling the floor out from under the wait-and-see crowd.

Microsoft, Salesforce, ServiceNow, and Google have all made similar moves in 2026. SAP’s reach across business processes and industries gives its direction particular weight for customers.

What This Means for Your Organization

If you run SAP, prepare for agent-native workflows inside your team’s daily tools. Start redesigning how decisions move, who owns what, and where an agent fits into the process.

If you use another vendor, examine its roadmap too. Agentic features are already shipping across operational software, giving your team a reason to prepare now.

Six months from now, the software your team relies on will work differently. If you spent this window preparing — redesigning workflows, testing agents in low-stakes processes, training your people — that investment pays off. If you spent it waiting for someone to tell you it was time, you are already behind.

SAP’s memo gave its workforce a directive to change how it works. Organizations running enterprise software in 2026 should use that signal to prepare their own teams for the changes arriving through their vendors.