# SAP: Knowledge graphs and governance are essential for enterprise AI agents Critical steps for companies transitioning from chatbots to autonomous AI agents. - **Source:** [VentureBeat](https://venturebeat.com) - **Published:** 2026-07-27 - **Topic:** [Data & AI](https://brifing.si/c/data-ai) - **Original article:** - **On Brifing:** https://brifing.si/s/ef874088-fa7a-40c3-b53d-0a1663966073 ## Transitioning from chatbots to autonomous agents Max McPhee, senior solution advisor at SAP, emphasized the importance of companies moving from chatbots to autonomous AI agents during his talk at VB Transform 2026. McPhee stated that these agents need to be grounded in the company's own context rather than general knowledge. This allows agents to behave more like a coworker than an assistant. SAP recommends using knowledge graphs and vector-embedded data to facilitate this transition. ## Building enterprise context with knowledge graphs McPhee explained that the principles used to onboard new employees also apply to AI agents, with adjustments for how software retrieves information. Knowledge graphs and vector-embedded data make it easier for agents to find and retrieve information. This approach helps companies overcome challenges like internal shorthand, ensuring smoother operations. ## Governance, identity, and security SAP offers a robust infrastructure for governance and process control in autonomous agents. McPhee highlighted that as a 50-year-old process company, SAP has modernized its governance to handle the flexibility of agents. Machine learning is regaining importance for validating agent behavior, with anomaly detection and ML-based validation acting as guardrails. ## Identity and permission management SAP's generative AI assistant, Joule, operates based on user access rights. Even if a user has permission to access S/4, Joule must also be provisioned for that access, eliminating the risk of bypassing controls. SAP ensures top-tier security with this model, preventing unauthorized access through agents. ## Balancing standard SAP with customized systems McPhee stressed the need to bridge SAP's knowledge with customer customizations and non-SAP systems. Many customers report that SAP constitutes only 10% of their landscape. Acquisitions like LeanIX and Signavio help map these non-SAP systems, enabling agents to understand enterprise interconnections. ## Automation investments SAP acquired Berlin-based automation company n8n, integrating it into Joule Studio, a low-code environment for building agents. McPhee warned that companies must modernize legacy systems to avoid performance issues, likening it to 'driving a Ferrari on a dirt track.' Upgrading infrastructure is essential for leveraging autonomous agents. ## The resurgence of machine learning McPhee noted that the rise of autonomous agents has brought machine learning back into focus. Anomaly detection and validation processes ensure agents operate securely. SAP has long used this approach for intelligent approval recommendations, enhancing reliability in business processes. ## The future of business processes SAP predicts that autonomous AI agents will transform business processes. Knowledge graphs, governance, and security measures are foundational to this shift. McPhee warned that without modernizing infrastructure, companies won't fully benefit from these technologies. SAP aims to guide customers through this transformation. ## Sources covering this story - https://venturebeat.com/orchestration/why-sap-says-enterprise-ai-agents-need-knowledge-graphs-and-governance — --- _Summary written by Brifing from the original reporting. Copyright in the original article belongs to its publisher; read the full piece at the link above._