Attack scenario · AWS identity & access

Excessive IAM permissions

Why are excessive AWS IAM permissions dangerous? Broad, wildcard permissions granted for convenience turn any compromise of a low-value workload into access to high-value systems.

Fictional, educational scenarioData exposurePrivilege takeoverLast updated · ATT&CK v20.1

Attack path · 4 stages

  1. 1

    Gain a foothold in a low-value workload

    T1190

  2. 2

    Use the instance role

    T1552.005

  3. 3

    Read secrets across the account

    T1555.006

  4. 4

    Access unrelated data stores

    T1530

For leadership

Executive summary

Most cloud breaches do not need a sophisticated exploit; they reuse permissions that were already granted. When a role used by a low-risk internal tool can also read every secret and bucket in the account, a bug in that tool becomes a company-wide data incident. Excessive permissions are invisible in day-to-day operations because nothing breaks — the risk only appears when an identity is misused. The practical goal is not perfection but removing the small number of high-impact permissions (identity administration, secrets, keys, sensitive data) from identities that do not need them.

For engineers

Technical summary

Policies with Action "*" or service-wide wildcards (s3:*, secretsmanager:*, kms:*) on Resource "*" grant capabilities far beyond the workload’s use. Any code-execution or credential-theft path against that workload inherits them.

Environment and assumptions

Fictional company: an internal reporting service runs on EC2 with an instance role carrying an inline policy of s3:*, secretsmanager:GetSecretValue and kms:Decrypt on all resources. The service only needs to read one reporting bucket.

  • The adversary gains code execution in the reporting service (for example through a vulnerable dependency).
  • IMDSv2 is enforced, but code running on the instance can still obtain role credentials.
  • Fictional and simplified for education.

Interactive

Attack sequence

Select a stage, or use the arrow keys, to see preconditions, evidence and controls.

Stage 1

Gain a foothold in a low-value workload

A vulnerability in the reporting service gives the adversary code execution on the instance.

Preconditions
  • Exploitable application vulnerability
Required access
  • Network access to the service.
Potential evidence
  • Application and WAF logs
  • Unusual outbound connections from the instance
Defensive controls
  • Patch and dependency management
  • Network restriction of internal tools
Text version of the attack sequence
  1. Gain a foothold in a low-value workload (T1190). A vulnerability in the reporting service gives the adversary code execution on the instance. Preconditions: Exploitable application vulnerability. Evidence: Application and WAF logs; Unusual outbound connections from the instance. Controls: Patch and dependency management; Network restriction of internal tools.
  2. Use the instance role (T1552.005). Code on the instance retrieves temporary role credentials from the metadata service. Preconditions: Code execution on the instance. Evidence: GuardDuty: instance credentials used from outside the instance’s expected context. Controls: Least-privilege instance roles; IMDSv2 (limits SSRF, not local code execution).
  3. Read secrets across the account (T1555.006). Because GetSecretValue is unscoped, the adversary reads database and third-party API credentials unrelated to reporting. Preconditions: secretsmanager:GetSecretValue on Resource "*"; kms:Decrypt on the secrets’ keys. Evidence: CloudTrail: GetSecretValue for secrets the role has never accessed before. Controls: Scope GetSecretValue to named secrets; Key policies that limit decrypt to expected consumers.
  4. Access unrelated data stores (T1530). With s3:* the adversary reads (or deletes) data in buckets the reporting service never used. Preconditions: s3:* on Resource "*". Evidence: S3 data events showing access to new buckets by the role. Controls: Resource-scoped S3 permissions; Bucket policies restricting sensitive buckets.

Technique mapping

MITRE ATT&CK and ATLAS techniques referenced in this scenario
IDTechniqueTactics (official)Stages
T1190Exploit Public-Facing ApplicationInitial Access1
T1552.005Unsecured Credentials: Cloud Instance Metadata APICredential Access2
T1555.006Credentials from Password Stores: Cloud Secrets Management StoresCredential Access3
T1530Data from Cloud StorageCollection4

Technique names and tactic assignments come from the imported MITRE data. Mapping a stage to a technique is Securitain’s analysis.

Why the attack can succeed

  • Permissions were sized for convenience, not for what the service does.
  • Secrets and keys were reachable by default rather than by explicit grant.
  • Nothing in normal operations reveals that a permission is unused.

What prevents or limits the attack

What may detect the attack

Detection signals and telemetry sources
SignalTelemetry sourceNotes
A role accessing a secret or bucket for the first timeCloudTrail (management + data events)—
Instance credentials used from outside the instanceGuardDuty—

Telemetry only helps if it is enabled before the incident and someone acts on it. Listing a signal here does not mean it is collected or monitored in your environment.

What Securitain may assess

  • Securitain can assess todayEffective-permission view that considers managed, inline, group, boundary and SCP policies together. Learn more
  • Securitain can assess todayDetection of identities with sensitive permissions such as broad S3 access, KMS decrypt or Secrets Manager reads. Learn more
  • Securitain can assess todayFinding-to-control mapping for supported frameworks (for example CIS, SOC 2, HIPAA, NIST and PCI DSS). Learn more

For leadership

Executive implications

confidentiality
Secrets for unrelated systems can be harvested, extending the incident beyond AWS.
integrity
s3:* includes delete and policy changes, enabling tampering.
compliance
Least privilege is an explicit expectation in most control frameworks.

No financial loss estimates are given; impact depends on the data, contracts and regulation involved.

Next steps

  1. List roles with wildcard access to IAM, Secrets Manager, KMS or S3.
  2. Start with roles attached to internet-facing or third-party-code workloads.
  3. Replace wildcards with resource-scoped grants based on last-accessed data.

References

MITRE ATT&CK® and ATT&CK® are registered trademarks of The MITRE Corporation. ATT&CK data © 2026 The MITRE Corporation. This work is reproduced and distributed with the permission of The MITRE Corporation (ATT&CK v20.1). Securitain is not affiliated with or endorsed by MITRE. Explanations, scenarios and control mappings on this page are Securitain’s own analysis. Sources and licensing

Stay in the loop

Cloud security insights, early access & compliance updates

Weekly InsightsEarly AccessExpert Content
No spam, everUnsubscribe anytimeWeekly updates
Chat with Securitain