// MITRE ATT&CK
T1698 · Indirect Prompt Injection
Adversaries may deliver malicious prompts to an LLM indirectly by embedding the prompts in content that the adversary expects may be processed by an LLM.(Citation: NIST indirect prompt injection) These prompts can be included in web pages,(Citation: Palo Alto Indirect Prompt Injection ITW) documents...
How to detect & mitigate it
Detecting Indirect Prompt Injection starts with your SIEM/EDR telemetry. Write a detection rule with the Sigma generator, analyze suspicious logs in the log analyzer, and place the technique on your coverage with the ATT&CK matrix.
Related techniques
T1053.005
Scheduled Task T1047
Windows Management Instrumentation T1129
Shared Modules T1059.007
JavaScript T1053.007
Container Orchestration Job T1559.002
Dynamic Data Exchange T1204.002
Malicious File T1053.003
Cron
Scheduled Task T1047
Windows Management Instrumentation T1129
Shared Modules T1059.007
JavaScript T1053.007
Container Orchestration Job T1559.002
Dynamic Data Exchange T1204.002
Malicious File T1053.003
Cron
Source: MITRE ATT&CK®. ATT&CK is a registered trademark of The MITRE Corporation. Content for educational purposes.