Published: May 14, 2026
Citation: IEEE Security & Privacy vol. 24, no. 3, (May-June 2026) pp. 52-58
Planning Note (06/09/2026):
NIST Mathematical Proof Supports Transition to a Continuous-Monitor-and-Update Security Model for AI Systems describes the significance of this new technical article from NIST's Apostol Vassilev.
Author(s)
Apostol Vassilev (NIST)
This article establishes information-theoretic limitations for robustness of artificial intelligence (AI) security and alignment. Knowing these limitations and preparing for their challenges is essential for responsible adoption of AI. Broader implications for cognitive reasoning limitations of AI systems are also proven.
This article establishes information-theoretic limitations for robustness of artificial intelligence (AI) security and alignment. Knowing these limitations and preparing for their challenges is essential for responsible adoption of AI. Broader implications for cognitive reasoning limitations of AI...
See full abstract
This article establishes information-theoretic limitations for robustness of artificial intelligence (AI) security and alignment. Knowing these limitations and preparing for their challenges is essential for responsible adoption of AI. Broader implications for cognitive reasoning limitations of AI systems are also proven.
Hide full abstract
Keywords
Artificial Intelligence; Weapons; Nuclear Weapons; Filtering; Filters; Circuits And Systems; Contacts; Radio Access Networks; Regional Area Networks; Videos; Video Equipment; Less Than Or Equal; Proof Of Theorem; Set Of Rules; Robust System; Finite Set; Broader Implications; Good And Bad; Intelligence Agencies; Binary String; Result Of Theorem; Human Language; Artificial Intelligence Systems; String Length; Cognitive Reasoning; Real Life Systems; Context Window; Adversarial Robustness; Information Theoretic Security; Chatbot
Control Families
None selected
