Record summary
A quick snapshot of what this page covers.
Techniques7Attack methods connected to this risk.
Mitigations18Defenses that may help with related attacks.
Domain2. Privacy & SecurityThe broad risk area this belongs to.
Risk profile
How this risk is described and categorized.
"General- purpose AI systems can cause or contribute to violations of user privacy. Violations can occur inadvertently during the training or usage of AI systems, for example through unauthorised processing of personal data or leaking health records used in training. But violations can also happen deliberately through the use of general- purpose AI by malicious actors; for example, if they use AI to infer private facts or violate security."
Domain2. Privacy & Security
Subdomain2.1 > Compromise of privacy by leaking or correctly inferring sensitive information
Entity2 - AI
Intent2 - Unintentional
Timing3 - Other
CategorySystemic risks
SubcategoryRisks to privacy
Suggested mitigations
Defenses that may help with related attacks.
Restrict Number of AI Model Queries
Business and Data UnderstandingDeployment+1 more
Control Access to AI Models and Data in Production
DeploymentMonitoring and Maintenance
AI Telemetry Logging
DeploymentMonitoring and Maintenance
Privileged AI Agent Permissions Configuration
Deployment
Single-User AI Agent Permissions Configuration
Deployment
AI Agent Tools Permissions Configuration
Deployment
Human In-the-Loop for AI Agent Actions
Deployment
Restrict AI Agent Tool Invocation on Untrusted Data
Deployment
Segmentation of AI Agent Components
DeploymentBusiness and Data Understanding
Input and Output Validation for AI Agent Components
Business and Data UnderstandingData Preparation+1 more
Restrict Library Loading
Deployment
Verify AI Artifacts
Business and Data UnderstandingData Preparation+1 more
Vulnerability Scanning
ML Model EngineeringData Preparation
User Training
Business and Data UnderstandingData Preparation+4 more
AI Bill of Materials
Business and Data UnderstandingData Preparation+1 more
Control Access to AI Models and Data at Rest
Business and Data UnderstandingData Preparation+2 more
Sanitize Training Data
Business and Data UnderstandingData Preparation+1 more
Maintain AI Dataset Provenance
Data PreparationBusiness and Data Understanding
Source
Research source for this risk, when available.
Included resource
International AI Safety Report 2025
Original source
MIT AI Risk Repository
Open the public repository used for AI risk records and taxonomy fields.