Epistemic Corruption Attacks: Poisoning What AI Security Systems Know Rather Than What They Do - Couverture souple

BHATNAGAR, PRANAV

Couverture souple

ISBN 10 : 9999339549 ISBN 13 : 9789999339544

Editeur : Eliva Press, 2026

Synopsis

Artificial intelligence systems are increasingly trusted to detect cyber threats, guide operational decisions, and interpret complex digital environments. Yet a deeper vulnerability is emerging one that does not attack what AI systems do, but what they believe. Modern AI systems continuously construct internal models of reality from the data they ingest stuff like logs, telemetry streams, behavioral signals, threat intelligence feeds, and external knowledge sources. These signals shape how machines interpret events, evaluate risks, and determine what actions to take. If those signals are manipulated, the system’s understanding of reality can quietly drift. Nothing appears broken. Dashboards look normal. Alerts still trigger. Confidence scores remain high. But beneath the surface, the machine’s beliefs have already changed. Epistemic Corruption Attacks explores how adversaries can manipulate the informational environment surrounding AI systems to influence their reasoning. Instead of attacking algorithms directly, attackers can reshape the signals that machines rely on to understand the world. Because the most dangerous failure in an AI system is not when it stops working. It is when it continues working with complete confidence while being completely wrong.

Détails de l'édition

• Éditeur: Eliva Press
• Date d'édition: 2026
• Langue: anglais
• ISBN 10: 9999339549
• ISBN 13: 9789999339544
• Reliure: Broché
• Nombre de pages: 289
• Coordonnées du fabricant: non disponible
• Personne responsable: gpsr@libri.de; gpsr@libri.de; ; Europaallee 1; Bad Hersfeld; 36244; Allemagne