Editorial illustration for SAP Brings Governance and Security to Enterprise AI Agents
SAP Adds Governance to Enterprise AI Agents
Most enterprise chatbots can answer questions. Few can actually run a business process without a human checking their work at every step. That gap is what SAP is trying to close, and at VB Transform 2026, Max McPhee, senior solution advisor at SAP, laid out why closing it has less to do with bigger models and more to do with what those models are allowed to know.
McPhee spoke with Rob Stretchay, lead analyst at VentureBeat Research, about the shift from assistant-style AI to agents that behave more like a colleague inside a company. His argument: general-purpose knowledge only gets an agent so far. The real unlock, if an enterprise wants something that operates like a coworker rather than a search box, is grounding that agent in the specific data, processes and internal language of the company running it.
SAP, whose customers run on decades of company-specific configurations and acronyms, has a direct stake in getting this right. McPhee framed the challenge as one of onboarding, not just engineering, and pointed to what that means for how enterprises structure their own data before an agent ever gets near it.
"When you are onboarding a new agent, I think it's important to acknowledge how you might onboard a new employee, but tune that for an agent," McPhee said. "The way that is really powerful is using knowledge graphs and having vector-embedded data, because that's a really easy format for an agent to be able to find and retrieve information."
Why this matters
SAP is betting that governance, not raw capability, becomes the bottleneck for enterprise AI agents, and that's a bet worth watching closely. McPhee's framing puts the company's decades of identity, access, and process controls to work on a new problem: agents that act on their own rather than wait for a human click. That's a real advantage over startups building agents from scratch, but it also means SAP's pitch depends on customers already trusting SAP with their core business data.
For developers and founders building agent products, the lesson is blunt: grounding an agent in company-specific context via knowledge graphs is table stakes, and governance can't be bolted on after the agent is already making decisions. Researchers should note the "emergent behavior" language McPhee uses. That's a tell that even SAP doesn't fully predict what these agents do once they're loose in a real workflow.
Enterprises will move slowly here, and rightly so. Worth watching: whether SAP's governance layer becomes a genuine standard other agent frameworks adopt, or just a moat SAP builds around its own ecosystem.
Related Reading
Business & Startups Nordic pilot adds Gemini for Education, NotebookLM to boost AI literacy
Business & Startups Kling launches Video O1, all-in-one model with MVL bridge using transformer
Business & Startups DeepSeek Seeks More Capital Weeks After USD 7B Funding Round
From Archives Monday.com joins 20 tech firms citing AI in workforce reductions
From Archives OpenAI's Micro keypad: A coder's tool that mystifies others
Common Questions Answered
What is the key difference between enterprise chatbots and SAP's enterprise AI agents?
Most enterprise chatbots can answer questions but require human verification at every step of a business process, whereas SAP's enterprise AI agents are designed to actually run business processes autonomously without constant human oversight. This capability gap is what SAP is working to close by focusing on governance and security rather than just model size.
How does SAP recommend onboarding new AI agents using knowledge graphs?
SAP recommends onboarding AI agents similarly to how you would onboard a new employee, but optimized for agents through the use of knowledge graphs and vector-embedded data. This approach makes it easier for agents to find and retrieve information in a format they can readily process and understand.
Why does SAP believe governance is more important than raw model capability for enterprise AI agents?
SAP is betting that governance, not raw capability, becomes the bottleneck for enterprise AI agents to operate effectively in business environments. By controlling what information agents are allowed to access and how they operate, organizations can enable autonomous agent action while maintaining security and compliance requirements.
What competitive advantage does SAP's decades of experience provide for enterprise AI agents?
SAP's extensive background in identity, access, and process controls gives it a significant advantage over startups building AI agents from scratch. The company can apply its established governance frameworks to the new problem of managing autonomous agents, rather than having to build these capabilities from the ground up.
Further Reading
- Shaping the Future of Secure AI Agents: How SAP and NVIDIA Co-Define Enterprise-Grade Agent Execution - SAP News Center
- Integration, Security, Ethics & Governance - SAP Architecture Learning
- SAP Business AI Platform | Build and Govern Agentic AI - SAP
- SAP Agent Hub: Observe and Govern Enterprise AI - SAP Community
- Govern and secure AI agents - Microsoft Learn
