Gate AI: LLM Security Benchmark Evaluation Methodology and Results

Published evaluations of prompt-injection and jailbreak detectors for Large Language Models often suffer from two systematic weaknesses: per-dataset threshold tuning and undisclosed operating points. We describe an evaluation harness that addresses both. The detector under evaluation is scored...

Code-Centric Detection of Vulnerability-Fixing Commits: A Unified Benchmark and Empirical Study

Automated detection of vulnerability-fixing commits VFCs is critical for timely security patch deployment, as advisory databases lag patch releases by a median of 25 days and many fixes never receive advisories. We present a comprehensive evaluation of code language model based VFC detection...

Hydra: Robust Hardware-Assisted Malware Detection

Malware detection using Hardware Performance Counters HPCs offers a promising, low-overhead approach for monitoring program behavior. However, a fundamental architectural constraint, that only a limited number of hardware events can be monitored concurrently, creates a significant bottleneck,...

Uncovering and Understanding FPR Manipulation Attack in Industrial IoT Networks

In the network security domain, due to practical issues -- including imbalanced data and heterogeneous legitimate network traffic -- adversarial attacks in machine learning-based NIDSs have been viewed as attack packets misclassified as benign. Due to this prevailing belief, the possibility of...

One Detector Fits All: Robust and Adaptive Detection of Malicious Packages from PyPI to Enterprises

The rise of supply chain attacks via malicious Python packages demands robust detection solutions. Current approaches, however, overlook two critical challenges: robustness against adversarial source code transformations and adaptability to the varying false positive rate FPR requirements of...

Analyzing PDFs like Binaries: Adversarially Robust PDF Malware Analysis Via Intermediate Representation and Language Model

Malicious PDF files have emerged as a persistent threat and become a popular attack vector in web-based attacks. While machine learning-based PDF malware classifiers have shown promise, these classifiers are often susceptible to adversarial attacks, undermining their reliability. To address this...

Hackers Exploit Legitimate Packer Software to Spread Malware Undetected

Threat actors are increasingly abusing legitimate and commercially available packer software such as BoxedApp to evade detection and distribute malware such as remote access trojans and information stealers. "The majority of the attributed malicious samples targeted financial institutions and...

Google Unveils RETVec - Gmail's New Defense Against Spam and Malicious Emails

Google has revealed a new multilingual text vectorizer called RETVec short for Resilient and Efficient Text Vectorizer to help detect potentially harmful content such as spam and malicious emails in Gmail. "RETVec is trained to be resilient against character-level manipulations including insertio...

APKHunt - Comprehensive Static Code Analysis Tool For Android Apps That Is Based On The OWASP MASVS Framework

APKHunt is a comprehensive static code analysis tool for Android apps that is based on the OWASP MASVS framework. Although APKHunt is intended primarily for mobile app developers and security testers, it can be used by anyone to identify and address potential security vulnerabilities in their cod...

How Carbon Black is Prioritizing Living Off the Land Attacks

What are Living Off the Land LoL Attacks? In recent years, Living off the Land Binaries and Scripts LOLBas have become increasingly popular tools for cybercriminals. These types of attacks leverage native, signed, and often pre-installed applications in malicious ways that their creators never...

Facebook Awards $100K to Researchers for Credential Spearphishing Detection Method

A group of researchers recently identified a real-time way to detect credential spearphishing attacks in enterprise settings. The discovery net the researchers $100,000 last week from Facebook, which awards money as part of its annual Internet Defense Prize partnership with USENIX Association. Th...