Cloud Security Fundamentals
Cloud security is identity and policy first, then reachable networks, resource policies, secrets, encryption and evidence.
Frame the problem
Security starts with a concrete asset, attacker capability and trust crossing.
Why the system fails
Broad policies, public resources and long-lived keys let one compromise act across unrelated services.
The important question is not “what is Cloud Security Fundamentals?” but “which assumption let untrusted data or an over-scoped identity cross cloud identity → policy → action → resource?” Trace the decision at the boundary, then constrain what can happen after the first control fails.
Design the control in layers
Start with the control closest to the interpretation or privilege boundary: Use least-privilege IAM and private-by-default resources Then add a control that reduces blast radius and telemetry that proves the decision was enforced.
The resulting design is not labelled secure. Record the identified controls, the known failure paths, the remaining exposure, and the evidence you would need during an incident.
| Prevent | Detect | Recover |
|---|---|---|
| Use least-privilege IAM and private-by-default resources · Prefer workload identity and short-lived credentials · Separate production accounts/projects and duties | Control-plane audit logs and unusual identity/resource combinations | Contain the affected identity or component, scope impact from audit evidence, and preserve a regression test. |
Key points
- Asset: Cloud data, control planes, workloads and billing capacity.
- Boundary: Cloud identity → policy → action → resource
- Primary control: Use least-privilege IAM and private-by-default resources
- Detection signal: Control-plane audit logs and unusual identity/resource combinations
- Always ask what limits damage when the primary control fails.
IAM Policy Boundary
Change the system and observe which assumption moves.
Follow the attack
Safe conceptual simulation: capability → missing control → crossed boundary → asset impact.
- 1Attacker starts with: A stolen human or workload identity, public resource caller or compromised CI runner.
- 2Broad policies, public resources and long-lived keys let one compromise act across unrelated services.
- 3The weak or missing boundary control is crossed: Cloud identity → policy → action → resource
- 4Impact: Data disclosure, destructive control-plane changes or expensive abuse.
- Data disclosure, destructive control-plane changes or expensive abuse.
Defend, detect, recover
One prevention is a single point of security failure. Layer it and make failure observable.
- • Use least-privilege IAM and private-by-default resources
- • Prefer workload identity and short-lived credentials
- • Separate production accounts/projects and duties
- • Control-plane audit logs and unusual identity/resource combinations
- • Contain the affected identity or component.
- • Scope access from audit evidence.
- • Fix the boundary and add a regression test.
- • Misconfiguration and new access paths can bypass the intended control.
- • A privileged insider or compromised control plane may still reach the asset.