Attack scenario · Data security

Sensitive data access through an overprivileged workload role

Can a compromised application read data it was never meant to access? A web application vulnerability gives an adversary the application’s AWS role — which can read far more data than the application uses.

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

Attack path · 4 stages

  1. 1

    Exploit the application

    T1190

  2. 2

    Obtain the workload role

    T1552.005T1528

  3. 3

    Enumerate data stores

    T1580T1619

  4. 4

    Read unrelated data

    T1530

For leadership

Executive summary

When an internet-facing application is compromised, the damage is defined by what that application’s cloud identity is allowed to do. Many applications run with permissions that include every table and bucket in the account because it was simpler at launch. That converts an ordinary application bug into a large data incident. Reducing each workload’s access to the specific data it serves is one of the highest-value, lowest-cost security investments available. It does not prevent the application bug, but it changes the outcome from "everything in the account was reachable" to "only this service’s own data was at risk" — a very different conversation with customers and regulators.

For engineers

Technical summary

Server-side request forgery against IMDSv1, or remote code execution in the container, yields the task or instance role credentials. Those credentials work wherever the role’s policies and the resource policies allow.

Environment and assumptions

Fictional company: a public customer portal runs on ECS. Its task role has dynamodb:* and s3:GetObject on all resources. The portal only uses one DynamoDB table and one bucket.

  • An application vulnerability exposes the task role credentials.
  • Fictional and simplified for education.

Interactive

Attack sequence

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

Text version of the attack sequence
  1. Exploit the application (T1190). A vulnerability allows command execution or SSRF in the portal. Preconditions: Exploitable vulnerability. Evidence: WAF / application logs. Controls: Patch management, WAF.
  2. Obtain the workload role (T1552.005, T1528). The adversary retrieves the task role credentials from the container credentials endpoint or metadata service. Preconditions: Code execution or SSRF to credential endpoints. Evidence: GuardDuty: credentials used outside the workload. Controls: IMDSv2; restrict SSRF egress; Least-privilege task roles.
  3. Enumerate data stores (T1580, T1619). The adversary lists tables and buckets reachable by the role. Preconditions: List permissions. Evidence: CloudTrail: ListTables, ListBuckets by the task role. Controls: Remove list permissions not used by the workload.
  4. Read unrelated data (T1530). Tables and buckets belonging to other systems are read. Preconditions: Wildcard resource permissions. Evidence: CloudTrail data events for DynamoDB / S3. Controls: Resource-scoped permissions; aws:SourceVpce conditions on sensitive stores.

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
T1528Steal Application Access TokenCredential Access2
T1580Cloud Infrastructure DiscoveryDiscovery3
T1619Cloud Storage Object DiscoveryDiscovery3
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

  • The workload role is broader than the workload.
  • Credentials obtained by SSRF or RCE work from anywhere unless network conditions are applied.

What prevents or limits the attack

What may detect the attack

Detection signals and telemetry sources
SignalTelemetry sourceNotes
Workload role used from outside its VPCGuardDuty; CloudTrail sourceIPAddress / vpcEndpointId—
Workload role accessing data stores it never used beforeCloudTrail data events—

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 todayDetection of identities with sensitive permissions such as broad S3 access, KMS decrypt or Secrets Manager reads. Learn more
  • Securitain can assess todayEffective-permission view that considers managed, inline, group, boundary and SCP policies together. Learn more
  • Securitain can assess todayPublic and cross-account exposure checks on S3, KMS, SQS, SNS and Secrets Manager resource policies. Learn more

For leadership

Executive implications

confidentiality
Data from unrelated systems can be exposed through one application bug.
compliance
Data minimization and least-privilege expectations are not met.

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

Next steps

  1. Start with internet-facing workloads.
  2. Replace wildcard resources with explicit ARNs.
  3. Add network conditions to sensitive stores.

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