Explore the whole discipline
Every boundary.
A path to deeper understanding.
Connect architecture, controls and investigation. Each topic brings together focused guidance, practical questions and relevant Sentinel queries.
25 topics
Core tracks
Security foundations
Map the system, establish the boundaries, and collect evidence that controls work.
Explore topicMicrosoft 365 Copilot
Control oversharing, understand protection boundaries, and investigate organizational data use.
Explore topicAzure AI & Foundry
Design identity, networking, secrets and policy around the AI workload you actually run.
Explore topicAgents & Copilot Studio
Constrain connectors, tool authority and the consequential actions an agent can take.
Explore topicDetection & response
Build useful telemetry, investigate competing explanations, and contain the relevant authority.
Explore topicGovernance & assurance
Make release decisions with named owners, reproducible evidence and explicit exceptions.
Explore topicIdentity & authorization
Trace delegated authority, token audiences, privileged access and federation trust.
Explore topicRetrieval & data security
Protect document permissions, ingestion integrity and the storage behind AI grounding.
Explore topicAdversarial testing
Design threat-led experiments, inspect real side effects and interpret evaluation limits.
Explore topicIdentity & access
Conditional Access
Separate user, service-principal and managed-identity coverage; investigate policy outcomes without assuming universal enforcement.
Explore topicPrivileged access
Review temporary elevation, role assignment paths and recovery access around AI administration.
Explore topicWorkload identities
Constrain federation, credential lifetime, resource scope and the behavior of nonhuman principals.
Explore topicApplication governance
Investigate consent, owners, credentials and configuration changes that alter application authority.
Explore topicData & retrieval
Retrieval authorization
Keep tenant and document access checks intact through search, vector retrieval, grounding and caches.
Explore topicKnowledge ingestion
Track source ownership, transformations and permission versions before content enters a serving index.
Explore topicStorage security
Distinguish Entra authorization, shared keys, SAS and anonymous access in the data plane.
Explore topicData protection
Apply sensitivity labels and DLP with explicit workload, content and licensing boundaries.
Explore topicAgents & application safety
Prompt injection
Treat prompts, retrieved documents and tool responses as untrusted inputs; constrain what the application can execute.
Explore topicMCP & tool security
Separate tool discovery, transport authentication, token audiences and backend authorization.
Explore topicAgent action safety
Bind approvals to exact actions and handle retries, partial completion and compensating operations deliberately.
Explore topicPlatform & operations
Secrets & key management
Separate secret access from vault administration; plan version changes, rotation, recovery and evidence.
Explore topicPrivate networking
Verify endpoint scope, DNS resolution, outbound dependencies and fallback paths in the actual execution environment.
Explore topicSupply-chain security
Bind release approval to source, dependencies, model configuration and deployable artifacts.
Explore topicDetection & evidence
Detection engineering
Design bounded hunts and analytics with explicit joins, identifiers, ingestion windows and validation cases.
Explore topicTelemetry quality
Verify source coverage and collection freshness while minimizing sensitive data in traces and investigations.
Explore topicNo topics match yet.
Try a product name, control or broader security term.