Artificial Intelligence & Machine Learning
,
Events
,
Next-Generation Technologies & Secure Development
7AI’s Lior Div on Building Knowledge Graphs, Human Oversight to Drive AI Accuracy
Artificial intelligence in security demands deep organizational context to produce accurate results. Without this context, AI may have the same skills as a new analyst who lacks tribal knowledge and struggles to interpret data. Teams must teach AI systems internal processes, historical patterns and investigative priorities to avoid poor outcomes, said Lior Div, co-founder and CEO of 7AI.
See Also: How Technical Debt Puts Critical Infrastructure at Risk
Organizations often begin with simple implementations, but success requires a structured knowledge graph that connects data, workflows and historical insights tied to specific security tasks. Security teams must extract expertise from analysts and embed it into AI defense tools, Div said.
“When AI is coming without knowing a specific organization, it is like somebody that joined the company without having this tribal knowledge. Once AI has the tribal knowledge, you will see improvements in the accuracy of what AI actually is doing for you,” he said.
In this video interview with Information Security Media Group at RSAC Conference 2026, Div also discussed:
- Why knowledge graphs improve AI-driven security outcomes;
- How human expertise shapes AI accuracy and trust;
- Steps to adopt AI through phased, use case-driven deployment.
Div leads AI-native cybersecurity company 7AI focused on detecting and responding to advanced threats. A serial entrepreneur, he previously co-founded Cybereason and is an expert in forensics, hacking, reverse engineering and encryption, with leadership experience from an elite Israeli cyber intelligence unit.

