Hyun Sim ↗
RESEARCH · 2025 · SSRN 5213825

A Unified Framework for Defending Against
AI-Driven APTs.

Hyun Sim · Vera Inari

01 / SUMMARY

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.

A/01

Threat intelligence integrity

Uses tamper-resistant sharing concepts to improve the reliability and provenance of collaborative threat intelligence across organizations.

A/02

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.

A/03

Privacy-preserving analytics

Considers encrypted-data analysis techniques so security analytics can operate with reduced exposure of sensitive underlying information.

A/04

AI behavior defense

Combines adversarial-ML resilience with behavioral auditing of AI systems to detect manipulation, evasion, and unsafe autonomous actions.

02 / CITED BY

Independent citation.

Crossref and OpenAlex record this work as cited by a peer-reviewed IEEE publication.

IEEE · 2026

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.

Citing paper ↗
03 / RESEARCH RECORD

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.

AUTHOR

Hyun Sim

Also known professionally as Kenneth Sim; cybersecurity, cloud security, network security, and AI security.

Profile ↗
IDENTIFIER

DOI 10.2139/ssrn.5213825

SSRN abstract 5213825 · posted 6 May 2025.

Resolve ↗
KEYWORDS

AI security · cybersecurity · zero trust

Advanced persistent threats, threat-intelligence sharing, privacy-preserving analytics, adversarial machine learning, and AI behavior auditing.

RESEARCH