محفوظ في:
التفاصيل البيبلوغرافية
المؤلف الرئيسي: Jaspers, Koen
التنسيق: Recurso digital
اللغة:الإنجليزية
منشور في: Zenodo 2026
الموضوعات:
الوصول للمادة أونلاين:https://doi.org/10.5281/zenodo.18189704
الوسوم: إضافة وسم
لا توجد وسوم, كن أول من يضع وسما على هذه التسجيلة!
جدول المحتويات:
  • <h2><strong>Abstract</strong></h2> <p>The emergence of AI-driven software agents introduces a new class of cybersecurity risk. Unlike traditional malware, AI-enabled threats adapt, self-modify, exploit ambiguity, and operate through legitimate interfaces rather than overt exploitation. Existing cybersecurity models—largely based on signature detection, perimeter defense, and binary trust assumptions—are insufficient against these behaviors.</p> <p>This paper introduces <strong>Diamond Integrity Architecture (DIA)</strong>: a structural cybersecurity framework designed for systems that must remain safe under adaptive, ambiguous, and AI-mediated conditions. Rather than attempting to identify malicious content or intent, DIA enforces <strong>identity preservation, phase-aware operation, and fail-closed behavior</strong> at the architectural level. The framework is domain-independent and applies to AI agents, automated systems, cloud services, and critical infrastructure.</p> <p>DIA reframes cybersecurity as a problem of <strong>structural coherence under change</strong>, rather than adversarial prediction. It is compatible with existing security tools and focuses on preventing unsafe action even when threats appear legitimate.</p>