A Unified Framework for Defending Against
AI-Driven APTs.
Hyun Sim · Vera Inari
Layered defense for AI-enabled persistent threats.
This paper examines how AI-assisted advanced persistent threats can automate reconnaissance, adaptation, evasion, and attack execution faster than traditional defensive workflows can respond.
Threat intelligence integrity
Uses tamper-resistant sharing concepts to improve the reliability and provenance of collaborative threat intelligence across organizations.
Zero-trust enforcement
Applies continuous verification and least-privilege access principles to reduce the ability of an intruder to move laterally after an initial foothold.
Privacy-preserving analytics
Considers encrypted-data analysis techniques so security analytics can operate with reduced exposure of sensitive underlying information.
AI behavior defense
Combines adversarial-ML resilience with behavioral auditing of AI systems to detect manipulation, evasion, and unsafe autonomous actions.
Independent citation.
Crossref and OpenAlex record this work as cited by a peer-reviewed IEEE publication.
AI²-Defense: An Autonomous Counter-AI Framework Against AI-Generated Threats in Consumer Systems
IEEE Transactions on Consumer Electronics, Vol. 72, No. 3, pp. 6851–6860. Crossref reference metadata cites DOI 10.2139/ssrn.5213825.
Publication metadata.
The authoritative publication record is maintained by SSRN. This page provides a stable first-party index and identity link for search engines and readers.
Hyun Sim
Also known professionally as Kenneth Sim; cybersecurity, cloud security, network security, and AI security.
DOI 10.2139/ssrn.5213825
SSRN abstract 5213825 · posted 6 May 2025.
AI security · cybersecurity · zero trust
Advanced persistent threats, threat-intelligence sharing, privacy-preserving analytics, adversarial machine learning, and AI behavior auditing.