IBI-000935 - Criticalpublie
Prompt Injection Bias
Biais de prompt injection
IBI Score
8.7
out of 10
Severity5_critique
Frequencyfrequent
Scope groups5
Detection difficulty4
🇫🇷 Définition (FR)
Manipulation d'un LLM par injection de texte malveillant dans le prompt.
🇬🇧 Definition (EN)
Manipulation of an LLM by injecting malicious text into the prompt.
TriAxis Taxonomy
Structural Classification
O
Origin
O2
Computational & Statistical
L
Lifecycle
L7
Deployment
S
Specific
S28
Interaction bias
💡
Concrete Example
Un utilisateur réussit à faire ignorer les instructions système à un LLM pour extraire des données confidentielles.
📊 Detection Metrics
Distribution ShiftDemographic Parity Difference
⚙ Alert Thresholds
Distribution Shift0.1
Demographic Parity Difference0.1
📚 Sources & Scientific References
- •Ignore Previous Prompt: Attack Techniques For Language ModelsFabio Perez, Ian Ribeiro (2022)
- •Not what you've signed up for: Compromising Real-World LLM-Integrated Applications with Indirect Prompt InjectionKai Greshake, Sahar Abdelnabi, Shailesh Mishra, Christoph Endres, Thorsten Holz, Mario Fritz (2023)
- •InjecAgent: Benchmarking Indirect Prompt Injections in Tool-Integrated Large Language Model AgentsQiusi Zhan, Zhixiang Liang, Zifan Ying, Daniel Kang (2024)
🛡 Impact & Mitigation
🎯 Vulnerable groups impacted
No specific groups.
🛠 Mitigation methodologies
- AUDIT_TIERS : Third-party Audit
🏢 Sectors concerned
Finance & Banking (Risk: /5)Media & Social Platforms (Risk: /5)
⚖ Applicable regulations
🌍 International
- •ISO_42001 : ISO/IEC 42001:2023
- •UNESCO_2021 : UNESCO - Recommandation sur l'éthique de l'IA
- •OECD_AI_PRINCIPLES : OCDE - Principes IA
- •COE_AI_CONVENTION_INT : Conseil de l'Europe - Convention IA
📋 Metadata
Contributor
Mohamadou SIDIBE
Institution
AI BIAIS INITIATIVE
Version
1
Created on
October 14, 2026
Updated on
November 4, 2026
UUID
6372e516-a34a-4b90-9e06-3d9b53fb7e8d
Explore other biases
250 documented biases, classified by the TriAxis taxonomy.