Minacce al sistema AI
Catalogo completo di vulnerabilità specifiche dell'IA e vettori di attacco
Minacce critiche
Immediate risks with potential for severe impact. Require urgent remediation.
Alto rischio
Serious vulnerabilities that should be addressed promptly to reduce exposure.
Strategie di difesa
Best practice e mitigazioni per ridurre l'esposizione alle minacce dell'IA.
Categorie di minacce
Prompt Injection
CriticalCrafted inputs designed to manipulate model behavior, override safeguards, or extract sensitive information.
Approccio ai test
- Craft adversarial prompts with hidden instructions or special characters
- Attempt multi-turn injection chaining
- Test for jailbreak bypass of alignment filters
- Evaluate output sanitization and safety layers
Training Data Poisoning
CriticalMalicious or biased data introduced into training pipelines, compromising model integrity and reliability.
Approccio ai test
- Analyze data provenance and supply chain
- Inject poisoned samples and assess downstream effects
- Test resilience to mislabeled or manipulated data
- Review validation and anomaly detection mechanisms
Model Inversion
HighReconstructing training data or sensitive attributes from model outputs, leading to privacy breaches.
Approccio ai test
- Attempt to recover representative training samples
- Test susceptibility to membership inference attacks
- Evaluate differential privacy protections
- Assess risk of leaking PII from embeddings
Adversarial Examples
HighInputs intentionally perturbed to cause misclassification, hallucinations, or other erroneous outputs.
Approccio ai test
- Generate gradient-based adversarial examples
- Apply noise and perturbation attacks
- Check model consistency across variations
- Evaluate robustness against transfer attacks
Model Stealing
HighExtraction of model functionality or parameters through repeated queries or side-channel analysis.
Approccio ai test
- Simulate query-based model extraction
- Analyze API rate limits and response variability
- Check for fingerprinting vulnerabilities
- Test throttling and monitoring protections
Data Memorization Leakage
HighSensitive information unintentionally memorized by AI models, retrievable via crafted prompts.
Approccio ai test
- Probe for known secret patterns in outputs
- Test for repeated exposure of sensitive training data
- Assess risk of accidental PII disclosure
Model Misuse & Malicious Automation
HighAI leveraged to perform tasks outside intended scope, enabling social engineering, spam, or automated attacks.
Approccio ai test
- Simulate misuse scenarios using sandbox models
- Test AI output moderation and guardrails
- Assess monitoring alerts for abnormal behaviors
Best practice per test e difesa
Ambiente di test sicuro
- Utilizza istanze sandbox o di replica per i test
- Non eseguire mai test non autorizzati sui sistemi di produzione
- Implementa funzionalità di monitoraggio, registrazione e rollback
Documentazione e osservabilità
- Mantenere prove e registri di test dettagliati
- Cattura le risposte del modello per la riproducibilità
- Taggare, classificare e organizzare casi di test per audit futuri
Conformità legale ed etica
- Rimanere nell'ambito e nei contratti autorizzati
- Rispetta la protezione dei dati, le leggi sulla privacy e la proprietà intellettuale
- Segui la divulgazione responsabile e la divulgazione coordinata delle vulnerabilità
Monitoraggio e mitigazione
- Implementa il rilevamento di anomalie per output AI insoliti
- Rivedere regolarmente limiti di velocità, accesso API e modelli di query
- Integra avvisi in tempo reale per minacce critiche