Skip to content

Disable Runtime Protection Service

Disable/Weaken Runtime Protection Service

Adversaries may attempt to degrade or fully disable protective mechanisms, including WAFs, RASP modules, and cloud monitoring agents. Tactics range from corrupting configuration files to forcibly terminating processes, exploiting vulnerabilities in the security agent itself, or downgrading version checks that block suspicious calls. By reducing the application’s protective layers, attackers can operate more freely and escalate subsequent activities.

Once runtime protection is weakened, other intrusion methods, such as injection or code execution, become far easier to execute without detection. If a cloud environment relies heavily on centralized security services or logging, severing these connections blinds defenders to ongoing malicious operations. This technique often appears mid-campaign, once attackers have enough privileges to interact with or reconfigure security infrastructure.

Mitigations

ID Mitigation Description
M1040 Behavior Prevention on Endpoint Deploy Cloud Application Detection & Response (CADR) product that baseline per-application API behavior and automatically quarantine tokens or workloads generating high-volume or atypical enumeration sequences.
M1026 Privileged Account Management Restrict who can invoke systemctl stop, sc stop, or cloud-agent API calls; require MFA + break-glass approval for any change to security services.
M1050 Exploit Protection Enable kernel-level exploit guards to block memory tampering that disables hooks or patches security engines.

Detection

ID Data Source Detection
DS0009 Process: Process Termination Alert when a security service binary is the target of kill, Stop-Service, or equivalent APIs—especially by non-signed binaries or unexpected users.
DS0015 Application Log Ingest AV/EDR/WAF watchdog logs; flag crashes, watchdog resets, or self-protection violations correlated with privilege-escalation activity.
DS0022 File: File Modification Detect on-disk tampering of security-agent binaries/configs (hash change, unsigned DLL overwrite) and raise high-severity alerts.