APromptRiskDBThreat intelligence atlas
AI Risk

Jailbreak in LLM Malicious Use - Backdoor Attack

"However, there are still ones who can leave holes in the training dataset, making LLMs appear safe on average, but generate harmful content under other specific conditions. This kind of attack can be categorized as "backdoor attack". Evan et al. developed a backdoor model that behaves as expected when trained, but exhibits different and potentially harmful behavior when deployed [81]. The results show that these...

AI Risk2. Privacy & Security2.2 > AI system security vulnerabilities and attacks1 - Pre-deployment

Record summary

A quick snapshot of what this page covers.

Techniques18Attack methods connected to this risk.
Mitigations25Defenses that may help with related attacks.
Domain2. Privacy & SecurityThe broad risk area this belongs to.

Risk profile

How this risk is described and categorized.

"However, there are still ones who can leave holes in the training dataset, making LLMs appear safe on average, but generate harmful content under other specific conditions. This kind of attack can be categorized as "backdoor attack". Evan et al. developed a backdoor model that behaves as expected when trained, but exhibits different and potentially harmful behavior when deployed [81]. The results show that these backdoor behaviors persist even after multiple security training techniques are applied."

Domain2. Privacy & Security
Subdomain2.2 > AI system security vulnerabilities and attacks
Entity1 - Human
Intent1 - Intentional
Timing1 - Pre-deployment
CategoryMalicious Use
SubcategoryJailbreak in LLM Malicious Use - Backdoor Attack

Suggested mitigations

Defenses that may help with related attacks.

Generative AI Guardrails

ML Model EngineeringML Model Evaluation+1 more
LifecycleML Model Engineering + 2 moreCategoryTechnical - ML

Generative AI Guidelines

ML Model EngineeringML Model Evaluation+1 more
LifecycleML Model Engineering + 2 moreCategoryTechnical - ML

AI Telemetry Logging

DeploymentMonitoring and Maintenance
LifecycleDeployment + 1 moreCategoryTechnical - Cyber

Sanitize Training Data

Business and Data UnderstandingData Preparation+1 more
LifecycleBusiness and Data Understanding + 2 moreCategoryTechnical - ML

Validate AI Model

ML Model EvaluationMonitoring and Maintenance
LifecycleML Model Evaluation + 1 moreCategoryTechnical - ML

AI Bill of Materials

Business and Data UnderstandingData Preparation+1 more
LifecycleBusiness and Data Understanding + 2 moreCategoryPolicy

Code Signing

Deployment
LifecycleDeploymentCategoryTechnical - Cyber

Vulnerability Scanning

ML Model EngineeringData Preparation
LifecycleML Model Engineering + 1 moreCategoryTechnical - Cyber

User Training

Business and Data UnderstandingData Preparation+4 more
LifecycleBusiness and Data Understanding + 5 moreCategoryPolicy

Model Hardening

Data PreparationML Model Engineering
LifecycleData Preparation + 1 moreCategoryTechnical - ML

Use Multi-Modal Sensors

Business and Data UnderstandingData Preparation+1 more
LifecycleBusiness and Data Understanding + 2 moreCategoryTechnical - Cyber

Input Restoration

Data PreparationML Model Evaluation+2 more
LifecycleData Preparation + 3 moreCategoryTechnical - ML

Deepfake Detection

DeploymentMonitoring and Maintenance+2 more
LifecycleDeployment + 3 moreCategoryTechnical - ML

Verify AI Artifacts

Business and Data UnderstandingData Preparation+1 more
LifecycleBusiness and Data Understanding + 2 moreCategoryTechnical - Cyber

Source

Research source for this risk, when available.