Attack Surface

An attack surface encompasses all potential entry points for malicious actors to compromise a system, ranging from software vulnerabilities in AI models (like LLMs and CNNs) to hardware weaknesses in autonomous vehicles and network communication channels in federated learning. Current research focuses on identifying and mitigating these vulnerabilities across diverse systems, employing techniques like adversarial training, moving target defenses, and the use of large language models for improved vulnerability assessment. Understanding and reducing attack surfaces is crucial for enhancing the security and reliability of increasingly interconnected systems, impacting fields from autonomous driving to AI safety and healthcare.

Papers

September 24, 2024

Cyber Knowledge Completion Using Large Language Models
Braden K Webb, Sumit Purohit, Rounak Meyur
Knowledge Graph Completion Cyber Physical System Knowledge Base Completion Attack Surface

September 2, 2024

An Investigation of Denial of Service Attacks on Autonomous Driving Software and Hardware in Operation
Tillmann Stübler, Andrea Amodei, Domenico Capriglione, Giuseppe Tomasso, Nicolas Bonnotte, Shawan Mohammed
Autonomous Vehicle Economic Operation False Sense Hardware Environment Autonomous Driving Software Attack Surface

July 8, 2024

If You Don't Understand It, Don't Use It: Eliminating Trojans with Filters Between Layers
Adriano Hernandez
Large Language Model Multi Layer Trojan Attack Residual Stream Max Filter Attack Surface Trojan Horse

May 31, 2024

Investigating and unmasking feature-level vulnerabilities of CNNs to adversarial perturbations
Davide Coppola, Hwee Kuan Lee
Adversarial Perturbation CNN Model Adversarial Framework Attack Surface

March 8, 2024

Defending Against Unforeseen Failure Modes with Latent Adversarial Training
Stephen Casper, Lennart Schulze, Oam Patel, Dylan Hadfield-Menell
Adversarial Attack Adversarial Training Failure Mode Novel Attack Attack Surface

February 21, 2024

Coercing LLMs to do and reveal (almost) anything
Jonas Geiping, Alex Stein, Manli Shu, Khalid Saifullah, Yuxin Wen, Tom Goldstein
Large Language Model Adversarial Attack New Attack Attack Surface

January 16, 2024

Security and Privacy Issues and Solutions in Federated Learning for Digital Healthcare
Hyejun Jeong, Tai-Myoung Chung
Solution Path Security Related Privacy Issue Digital Healthcare Attack Surface

December 30, 2022

RL and Fingerprinting to Select Moving Target Defense Mechanisms for Zero-day Attacks in IoT
Alberto Huertas Celdrán, Pedro Miguel Sánchez Sánchez, Jan von der Assen, Timo Schenk, Gérôme Bovet, Gregorio Martínez Pérez, Burkhard Stiller
Reinforcement Learning Human Like RL Zero Day Fingerprinting Lemma Attack Surface Moving Target Defense Heterogeneous Zero Day Attack

September 28, 2022

A Closer Look at Evaluating the Bit-Flip Attack Against Deep Neural Networks
Kevin Hector, Mathieu Dumont, Pierre-Alain Moellic, Jean-Max Dutertre
Deep Neural Network Adversarial Learning Glance Annotation Attack Vector Attack Surface Parameter Based Attack

August 27, 2022

Network-Level Adversaries in Federated Learning
Giorgio Severi, Matthew Jagielski, Gökberk Yar, Yuxuan Wang, Alina Oprea, Cristina Nita-Rotaru
Attack Surface Client Side

Attack Surface

Papers

Cyber Knowledge Completion Using Large Language Models

An Investigation of Denial of Service Attacks on Autonomous Driving Software and Hardware in Operation

If You Don't Understand It, Don't Use It: Eliminating Trojans with Filters Between Layers

Investigating and unmasking feature-level vulnerabilities of CNNs to adversarial perturbations

Defending Against Unforeseen Failure Modes with Latent Adversarial Training

Coercing LLMs to do and reveal (almost) anything

Security and Privacy Issues and Solutions in Federated Learning for Digital Healthcare

RL and Fingerprinting to Select Moving Target Defense Mechanisms for Zero-day Attacks in IoT

A Closer Look at Evaluating the Bit-Flip Attack Against Deep Neural Networks

Network-Level Adversaries in Federated Learning