Privilege escalation happens when a user or process gains access rights beyond what was intended. On secure platforms, every permission, every role, should map precisely to a function. When that control breaks—whether through a misconfigured API, vulnerable service, or unpatched kernel—the attacker moves upward. They bypass intended limits. They hijack the platform’s trust model.
There are two main paths: vertical and horizontal. Vertical privilege escalation jumps from lower to higher access levels, reaching admin or root. Horizontal escalation happens when a user gains access to another user’s data or functions without increasing privilege level. Both methods exploit weak boundaries in platform security architecture.
Common causes include insecure role assignments, overly broad IAM policies, flawed session handling, and chaining of minor vulnerabilities. Sometimes the escalation hinges on logic errors—misapplied checks that trust the wrong condition. In containerized environments, privilege escalation can occur when isolation fails or a process gains dangerous host-level privileges via misconfigured mounts or escalated capabilities.