SAP API Policy

SAP API Policy and AI employees: what is allowed.

Since April 2026, SAP has regulated how AI systems may use SAP interfaces. What this means for AI employees.

Short answer

AI employees may continue to work with SAP as long as the AI does not decide on its own which SAP interfaces to call. At bee-hub.ai, data exchange with SAP runs through fixed, programmed workflows; the AI works on the data delivered, and your team posts.

What section 2.2.2 prohibits.

Section 2.2.2 of the SAP API Policy (version 4.2026a) prohibits using SAP APIs to interact or integrate with (semi-)autonomous or generative AI systems that plan, select or execute sequences of API calls. SAP-endorsed architectures and pathways are excepted. Systematic or large-scale data extraction is also prohibited.

What is meant is an AI that decides on its own which SAP interface to call, when and in what order.

What remains allowed.

  • Fixed workflows: SAP's FAQ on the policy explicitly excludes fixed call sequences without AI decision-making (question 40).
  • Outbound interfaces: According to the FAQ, data that SAP itself sends out through its interfaces is not affected either (question 12).

How bee-hub.ai's AI employees work.

  • Data exchange with SAP runs through fixed, programmed workflows, not through decisions by the AI.
  • The AI works on the data these workflows deliver: it reads documents, checks and prepares.
  • The AI never decides on an SAP call itself.
  • It places documents in SAP as parked documents, and your team posts them.

This is how our document processing with SAP works in live operation today.

Does this also apply to SAP Business One?

The policy neither names SAP Business One nor excludes it. Our setup is compliant in both cases: fixed workflows, no AI decisions about SAP calls, posting by your team.

SAP and AI employees on your own case.

30 minutes, one process from your business.

Book a demo
Source: SAP SE, SAP API Policy, version 4.2026a, section 2.2.2: help.sap.com