Archive
161 articles, chronologically ordered.
2026
AI Security Digest — July 16, 2026News & Trends
Antiproof: Synthesizing Vulnerability Detectors and Proofs of ExploitabilityPaper Review
AutoTrace: From Patches to Triggers via Agentic Interprocedural ExplorationPaper Review
Bulkhead: Automated Semantic Detection and Remediation of Container Escape VulnerabilitiesPaper Review
Cross-Cutting Security Analysis of LLM-Generated Code via Metamorphic Testing and Association Rule MiningPaper Review
Not All Agent Topologies Leak Equally: Structure, Language, and Privacy in LLM Multi-Agent SystemsResearch Paper
Open-Source Intelligence for Code Provenance and the Security Patterns that Separate Human and Large-Language-Model Implementations of Common Programming TasksPaper Review
PVDetector: Detecting Prompt Injection Attacks on Purpose-Specific LLM Agents through Policy-Violation Concept AnalysisPaper Review
Silent Alarm: A J-Space Protocol for Comparing Danger Recognition Across Models and Quantization LevelsPaper Review
Skills That Don't Exist: A Large-Scale Study of Hallucinated Skill Recommendation in LLM AgentsPaper Review
Stability Buys Time: A Re-Keying Game for Encrypted Multi-Agent ControlPaper Review
TrustVLA: Mechanism-Guided Inference-Time Defense Against Vision-Language-Action BackdoorsPaper Review
Watermark Forensics for Generative Models: An Information-Theoretic PerspectivePaper Review
When Binaries Talk Back: Representation-Confusion Attacks on LLM-Assisted Reverse EngineeringPaper Review
AdvNav: Behavior-Guided Black-Box Adversarial Attacks on Vision-Language NavigationPaper Review
Agent Hacks Agent: Autoresearch for Production-Agent Red-TeamingPaper Review
AI Security Digest — July 15, 2026News & Trends
AMT-X: Phase-Structured Multi-Turn Red-Teaming with Checklist-Gated EvaluationPaper Review
Cross-Layer Misalignment Detection in Agent Skills: A Progressive Loading-Aware Contrastive Learning ApproachPaper Review
Distributed Denial of Science: How Indirect Data Poisoning of AI Systems Can Industrialize Scientific FraudPaper Review
LLM-Guided Program Evolution for Targeted Black-Box Attacks on Perceptual Hash AlgorithmsPaper Review
Mako: A Self-Evolving Agentic Operating System (SE-AOS) for Autonomous Web ExploitationPaper Review
NetInjectBench: Benchmarking Indirect Prompt Injection in Tool-Using Large Language Model Agents for Network OperationsPaper Review
One Token Is Enough: Fingerprinting and Verifying Large Language Models from Single-Token Output DistributionsPaper Review
Understanding Implicit Trust Errors in Core Carrier Networks through Multi-Agent Flaw Discovery and AnalysisPaper Review
When cheap gradients fail: the measurement cost of attacking quantum classifiersPaper Review
When Local Monitors Miss Compositional Harm: Diagnosing Distributed Backdoors in Multi-Agent SystemsPaper Review
AI Security Digest — July 14, 2026News & Trends
Federated Learning Architecture: Data Privacy and System Security ApproachesPaper Review
OpenLongTail: Generative Scaling of Long-Tail Driving DataPaper Review
Optimizing Against Safety Representations: Activation-Guided Adversarial Suffixes and the Geometry of RefusalPaper Review
Statistically Undetectable Backdoors in Deep Neural NetworksPaper Review
Triggering Stealthy Feature Map Backdoors via Physical Fault Injection in Embedded Neural NetworksPaper Review
VEXAIoT: Autonomous IoT Vulnerability EXploitation using AI AgentsPaper Review
When Routes Run Out: Adversarial Co-Learning and Explainable Robustness in Quantum Repeater NetworksPaper Review
AI Security Digest — July 13, 2026News & Trends
Just Ask the Model: Natural Language Autoencoders and the Legible MindPaper Walkthrough
AI Security Digest — July 12, 2026News & Trends
This Week in AI Security — July 12, 2026News & Trends
AI Security Digest — July 11, 2026News & Trends
EdgeRefine: Privacy-Utility Balance for Graphs via Jaccard Sampling under Edge Differential PrivacyPaper Review
Efficient Safety Alignment of Language Models via Latent Personality TraitsPaper Review
Functional and Secure Code Generation with Task VectorsPaper Review
Open Models, Open Risks: Measuring Unsafe Generation in Text-to-Image Models In the WildPaper Review
Out of Sight: Compression-Aware Content Protection against Agentic CrawlersPaper Review
Persuasion Attacks Can Decrease Effectiveness of CoT MonitoringPaper Review
Prismata: Confining Cross-Site Prompt Injection in Web AgentsPaper Review
ScopeJudge: Cost-Aware Pre-Execution Gating for Offensive Security AgentsPaper Review
Securing Autonomous Vehicle Systems via Twin-Aware Federated Reinforcement LearningPaper Review
Token-Flow Firewall: Semantic Runtime Auditing for Persistent AI AgentsPaper Review
Safety Alignment Should Be Made More Than Just a Few Tokens DeepResearch Paper
MM-Snowball: Evaluating and Mitigating Hallucination Snowballing in Multimodal Multi-Turn DialoguePaper Review
The Invitation Trap: Proactive Availability Backdoor in LLMs via Conversational InductionPaper Review
BadBone: Backdoor Attacks Against Backbone Models in Visual Prompt LearningPaper Review
From Prompt Injection to Persistent Control: Defending Agentic Harness Against Trojan BackdoorsPaper Review
AI Security Digest — May 31, 2026News & Trends
This Week in AI Security — May 31, 2026News & Trends
A Bayesian Approach to Membership Inference for Statistical ReleasePaper Review
AI Security Digest — May 30, 2026News & Trends
AliMark: Enhancing Robustness of Sentence-Level Watermarking Against Text ParaphrasingPaper Review
Audio Jailbreaks in Large Audio-Language Models: Taxonomy, Attack-Defense Analysis, and Cost-Aware EvaluationPaper Review
Beyond English and Evasion: A Human-Annotated Multi-Domain Benchmark for High-Stakes LLM Safety Evaluation in ChinesePaper Review
Hidden in Plain Tokens: Simply Robust, Gradient-Free Watermark for Synthetic AudioPaper Review
Hijacking Agent Memory: Stealthy Trojan Attacks Through Conversational InteractionPaper Review
Minimal Prompt Perturbations Lead to Code Vulnerabilities: Prompt Fragility and Hidden-State Signals in Coding LLMsPaper Review
Reasoning as an Attack Surface: Adaptive Evolutionary CoT Jailbreaks for LLMsPaper Review
Relevance as a Vulnerability: How Web Retrieval Degrades Safety Alignment in LLM AgentsPaper Review
Robust LLM Watermarking with Minimal Semantic Distortion for IP ProtectionPaper Review
SafeCtrl-RL: Inference-Time Adaptive Behaviour Control for LLM Dialogue via RL-Driven Prompt OptimisationPaper Review
Summoning the Oracle to Slay It: Mitigating Look-Ahead Bias in Financial Backtesting with Large Language ModelsPaper Review
AI Security Digest — May 28, 2026News & Trends
Cordyceps: Covert Control Attacks on LLMs via Data PoisoningPaper Review
Ellipsoid Control: A White-list Jailbreak Defense via Benign Latent ModelingPaper Review
Five Queries Are Enough: Query-Efficient and Surrogate-Free Membership Inference Attacks on RAG via EntailmentPaper Review
IterInject: Indirect Prompt Injection Against LLM Agents via Feedback-Guided Iterative OptimizationPaper Review
Jailbreak to Protect: Buffering and Reinforcing via Temporary Jailbreaking for Safe Fine-Tuning in Large Language ModelsPaper Review
Localization then Neutralization: Gradient-guided Token Suppression against Visual Prompt Injection AttackPaper Review
Poisoning the Watchtower: Prompt Injection Attacks Against LLM-Augmented Security Operations Through Adversarial Log ContentPaper Review
Prompt Injection Detection is Regime-Dependent: A Deployment-Aware Evaluation with Interpretable Structural SignalsPaper Review
Resolving the Correct Library: A Loader-Level Defense Solution Against Shared Object HijackingPaper Review
Steering Beyond the Support: Adversarial Training on Unsupervised Jailbroken Activation SimulationPaper Review
Test-Time Training Undermines Safety GuardrailsPaper Review
AI Security Digest — May 27, 2026News & Trends
An Efficient and Privacy-Preserving Architecture for Cross-Institutional Collaborative RAGPaper Review
APT-Agent: Automated Penetration Testing using Large Language ModelsPaper Review
Building an Adversarial Malware Dataset by Family and Type: Generation, Evasion, and Poisoning EvaluationPaper Review
Evo-Attacker: Memory-Augmented Reinforcement Learning for Long-Horizon Tool Attacks on LLM-MASPaper Review
How Agentic AI Coding Assistants Become the Attacker's ShellPaper Review
LLM-as-a-Reviewer: Benchmarking Their Ability, Divergence, and Prompt Injection Resistance as Paper ReviewersPaper Review
QML-PipeGuard: Drift-Aware Behavioral Fingerprinting for Quantum Machine Learning Pipeline IntegrityPaper Review
Referential Security as a New Paradigm for AI EvaluationsPaper Review
SAMark: A Self-Anchored Text Watermarking with Paragraph-Level Paraphrase RobustnessPaper Review
Security in the Fine-Tuning Lifecycle of Large Language Models: Threats, Defenses,Evaluation, and Future DirectionsPaper Review
Semantic Validation of Packer Identification Tools: Characterization, Repair, and Downstream ImpactPaper Review
When Interpretability Becomes a Liability: Adversarial Attacks on CBM Concept LayersPaper Review
AI Security Digest — May 24, 2026News & Trends
This Week in AI Security — May 24, 2026News & Trends
AI Security Digest — May 23, 2026News & Trends
AI Security Digest — May 21, 2026News & Trends
AI Security Digest — May 20, 2026News & Trends
AI Security Digest — May 18, 2026News & Trends
AI Security Digest — May 17, 2026News & Trends
This Week in AI Security — May 17, 2026News & Trends
AI Security Digest — May 10, 2026News & Trends
This Week in AI Security — May 10, 2026News & Trends
AI Security Digest — May 09, 2026News & Trends
AI Security Digest — May 07, 2026News & Trends
AI Agent Traps: When the Environment Becomes the AttackerPaper Walkthrough
AI Security Digest — April 22, 2026News & Trends
AI Security Digest — April 21, 2026News & Trends
AI Security Digest — April 20, 2026News & Trends
AI Security Digest — April 19, 2026News & Trends
This Week in AI Security — April 19, 2026News & Trends
AI Security Digest — April 18, 2026News & Trends
Security of Autonomous AI Agents: Trust Boundary-Based Attack Surface Analysis and TrendsResearch Paper
AI Security Digest — April 13, 2026News & Trends
AI Security Digest — April 12, 2026News & Trends
This Week in AI Security — April 12, 2026News & Trends
AI Security Digest — April 11, 2026News & Trends
AI Security Digest — April 10, 2026News & Trends
AI Security Digest — April 07, 2026News & Trends
NeuroStrike: Neuron-Level Attacks on Aligned LLMsPaper Walkthrough
AI Security Digest — April 06, 2026News & Trends
AI Security Digest — April 05, 2026News & Trends
This Week in AI Security — April 05, 2026News & Trends
AI Security Digest — April 04, 2026News & Trends
AI Security Digest — April 03, 2026News & Trends
AI Security Digest — April 02, 2026News & Trends
AI Security Digest — April 01, 2026News & Trends
AI Security Digest — March 31, 2026News & Trends
AI Security Digest — March 30, 2026News & Trends
AI Security Digest — March 29, 2026News & Trends
AI Security Digest — March 28, 2026News & Trends
Bridging Models and Agents: Protocol Architectures and Security in MCP & A2AResearch Paper
Fool Me If You Can: On the Robustness of Binary Code Similarity Detection ModelsResearch Paper
LLM-Based Drug Term Detection in Korean Messenger ConversationsResearch Paper
Trends in Attacks and Defenses against Retrieval-Augmented Generation (RAG) SystemsResearch Paper
Analysis of Watermarking for AI-Generated TextResearch Paper
The Art of English Idioms: Why We Kick Buckets, Break Legs, and Beat Around BushesTutorial
Idioms: Turbo-Charging Neural Decompilation with User-Defined TypesPaper Review
Getting Started with Dezoomify Plus: A Practical Intro GuideProject
Visualizing RAG Security: A Deep Dive with RAG-Vis PlaygroundProject
Pickle Deserialization Attacks: Understanding Python's Silent RCE VulnerabilityTutorial
Pickleguard: Defending Python Applications Against Pickle Deserialization AttacksProject
MOEVIL: Poisoning Experts to Compromise the Safety of Mixture-of-Experts LLMsPaper Walkthrough
2025
LLM Red-Teaming: A Survey of Attack Strategies and Defense MechanismsTutorial
GASLITE: Poisoning Dense Embedding-based Retrieval SystemsPaper Walkthrough
Rescuing the Unpoisoned: Efficient Defense against Knowledge Corruption Attacks on RAG SystemsResearch Paper
AgentFuzz: Automatic Detection of Taint-Style Vulnerabilities in LLM-based AgentsPaper Walkthrough
An Information Theoretic Approach to Machine UnlearningPaper Walkthrough
Contrastive Learning for Code Vulnerability Type ClassificationPaper Walkthrough
Machine Unlearning for LLMs: Foundations and the AltPO ApproachPaper Walkthrough
Membership Inference Attacks on Retrieval-Augmented Generation: A Comprehensive SurveyPaper Walkthrough
AMIDES: Adaptive Misuse Detection for SIEM Rule EvasionsPaper Walkthrough
Instruction Backdoor Attacks against Customized LLMsPaper Walkthrough
2024
AutoDAN: Generating Stealthy Jailbreak Prompts on Aligned Large Language ModelsPaper Walkthrough
Teach LLMs to Phish: Stealing Private Information from Language ModelsPaper Walkthrough
Friends Don't Let Friends Make Bad Graphs: A Data Visualization GuideTutorial
Foice: Attacking Voice Authentication Systems with a Single Face ImagePaper Walkthrough
LINT: On Large Language Models' Resilience to Coercive InterrogationPaper Walkthrough
PassBERT: Improving Real-world Password Guessing via Bi-directional TransformersPaper Walkthrough