MITRE ATT&CK technique · T1685.002
Disable or Modify Tools: Disable or Modify Cloud Log
How can CloudTrail logging be disabled or modified by an attacker? Adversaries with sufficient permissions stop, delete or reconfigure cloud audit logging — or tamper with where logs are delivered — to limit evidence of their actions. In ATT&CK this technique moved to the Defense Impairment tactic in v19 and replaced T1562.008.
For leadership
Executive takeaway
How it works in AWS
- StopLogging or DeleteTrail removes recording entirely.
- PutEventSelectors or UpdateTrail quietly narrows what is recorded (for example dropping write events or other regions).
- Changing the log bucket policy or lifecycle removes delivered evidence.
Sub-technique of T1685 Disable or Modify Tools.
What the attacker needs
- cloudtrail:* administrative permissions on the trail
- No SCP protecting the trail
Relevant AWS API calls
| API / event | Why it matters |
|---|---|
| cloudtrail:StopLogging | Stops recording; the call itself is logged. |
| cloudtrail:PutEventSelectors | Can exclude management write events or data events. |
| cloudtrail:UpdateTrail | Can disable multi-region or change destinations. |
| cloudtrail:DeleteTrail | Removes the trail. |
A risky permission is not an observed attack
Per-account trails editable by account admins; logs stored in the same account; no SCP protection.
StopLogging / PutEventSelectors / UpdateTrail events, GuardDuty logging-disabled findings, gaps in delivery.
Preventive and limiting controls
- Organization trail managed centrally
- SCP denying trail modification in member accounts
- Separate log-archive account with Object Lock and integrity validation
Mapped cloud controls
- preventiveService control policy guardrails
An SCP denying cloudtrail:StopLogging, DeleteTrail, UpdateTrail and PutEventSelectors on the organization trail blocks member-account administrators from tampering with it.
- restrictiveOrganization-wide, tamper-resistant CloudTrail
Delivering logs to a separate log-archive account with integrity validation means a compromised workload account cannot delete or silently edit delivered logs.
- detectiveAlerting on identity and logging changes
StopLogging, DeleteTrail, UpdateTrail and PutEventSelectors are themselves recorded and should page someone.
- detectiveManaged threat detection across accounts and regions
GuardDuty has a specific finding for CloudTrail logging being disabled (Stealth:IAMUser/CloudTrailLoggingDisabled) and anomaly-based findings for other unusual defense-impairment API calls.
See assumptions and validation evidence in the Cloud Control Explorer →
Detection signals and telemetry
- Trail configuration changes — CloudTrail (EventBridge rule for real-time alerting)
- Logging disabled findings — GuardDuty
MITRE detection strategies (official, cloud-relevant analytics)
- DET0289 Detection Strategy for Disable or Modify Cloud Log
- AN0801 (IaaS): Cloud API events where logging services are stopped, deleted, or modified in a way that disables audit visibility. Defender view: unauthorized StopLogging, DeleteTrail, or UpdateSink operations correlated with privileged user activity.Log sources: AWS:CloudTrail — Stop logging for an existing CloudTrail; gcp:config — UpdateSink request modifying log export destinations
- AN0802 (Identity Provider): Disabling or modifying sign-in or audit log collection for user activities. Defender view: policy or configuration updates removing logging coverage for critical accounts.Log sources: azure:policy — DisableAuditLogs or ConditionalAccess logging changes
- AN0803 (Office Suite): Disabling mailbox or tenant-level audit logging, often using Set-MailboxAuditBypassAssociation or downgrading license tiers. Defender view: sudden absence of mailbox activity logging for monitored users.Log sources: m365:unified — Set-MailboxAuditBypassAssociation or disabling Advanced Auditing
- AN0804 (SaaS): Disabling or altering security and audit logs in SaaS admin panels (e.g., Slack, Zoom, Salesforce). Defender view: API calls or admin console changes that stop event exports or logging integrations.Log sources: saas:audit — Log export integration removed or disabled
MITRE mitigations
Official MITRE description (excerpt)
An adversary may disable or modify cloud logging capabilities and integrations to limit what data is collected on their activities and avoid detection. Cloud environments allow for collection and analysis of audit and application logs that provide insight into what activities a user does within the environment. If an adversary has sufficient permissions, they can disable or modify logging to avoid detection of their activities.
Common misconceptions
- “An attacker who stops CloudTrail leaves no trace.”
- The StopLogging call is recorded before logging stops — if alerts are routed and owned.
Attack scenarios using this technique
Privilege escalation through IAM policy relationships
CloudTrail logging impairment
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