Record summary
A quick snapshot of what this page covers.
Techniques1Attack methods connected to this risk.
Mitigations0Defenses that may help with related attacks.
Domain2. Privacy & SecurityThe broad risk area this belongs to.
Risk profile
How this risk is described and categorized.
Domain2. Privacy & Security
Subdomain2.2 > AI system security vulnerabilities and attacks
Entity3 - Other
Intent2 - Unintentional
Timing1 - Pre-deployment
CategorySoftware Security Issues
SubcategoryProgramming Language
Suggested mitigations
Defenses that may help with related attacks.
No propagated mitigations. No defense is available through the connected attack methods.
Source
Research source for this risk, when available.
Included resource
Risk Taxonomy, Mitigation, and Assessment Benchmarks of Large Language Model Systems
Original source
MIT AI Risk Repository
Open the public repository used for AI risk records and taxonomy fields.