Optimizing Canaries for Privacy Auditing with Metagradient Descent

In this work we study black-box privacy auditing, where the goal is to lower bound the privacy parameter of a differentially private learning algorithm using only the algorithm's outputs i.e., final trained model. For DP-SGD the most successful method for training differentially private deep...

Attacking Interpretable NLP Systems

Studies have shown that machine learning systems are vulnerable to adversarial examples in theory and practice. Where previous attacks have focused mainly on visual models that exploit the difference between human and machine perception, text-based models have also fallen victim to these attacks...

Microsoft at Black Hat USA 2025: A unified approach to modern cyber defense

Microsoft will be at Black Hat USA 2025, August 5–7 in Las Vegas, and we’re bringing you a unified, practitioner-driven experience built around real-world insights, threat intelligence, incident response, and hands-on AI expertise. We believe security teams are strongest when intelligence, tools,...

Breaking the Illusion of Security Via Interpretation: Interpretable Vision Transformer Systems under Attack

Vision transformer ViT models, when coupled with interpretation models, are regarded as secure and challenging to deceive, making them well-suited for security-critical domains such as medical applications, autonomous vehicles, drones, and robotics. However, successful attacks on these systems ca...

PT-2025-31065

Name of the Vulnerable Software and Affected Versions Linux kernel affected versions not specified Description A flaw exists in the Linux kernel’s networking scheduler net/sched related to handling empty red-black trees within the htb lookup leaf function. Specifically, a BUG ON condition can be...

Key Takeaways from the Take Command Summit 2025: Inside the SOC – Expert Stories from the Frontlines of Threat Hunting and Malware Detection

What does it really look like to detect, contain, and respond to modern cyber threats in real time? At the Take Command 2025 Virtual Cybersecurity Summit, Inside the SOC session offered a behind-the-scenes look at how security teams are tackling everything from ransomware staging to advanced soci...

PLA: Prompt Learning Attack against Text-To-Image Generative Models

Text-to-Image T2I models have gained widespread adoption across various applications. Despite the success, the potential misuse of T2I models poses significant risks of generating Not-Safe-For-Work NSFW content. To investigate the vulnerability of T2I models, this paper delves into adversarial...

Mitigating Watermark Stealing Attacks in Generative Models Via Multi-Key Watermarking

Watermarking offers a promising solution for GenAI providers to establish the provenance of their generated content. A watermark is a hidden signal embedded in the generated content, whose presence can later be verified using a secret watermarking key. A threat to GenAI providers are \emphwaterma...

SUSE CVE-2025-38177

In the Linux kernel, the following vulnerability has been resolved: schhfsc: make hfscqlennotify idempotent hfscqlennotify is not idempotent either and not friendly to its callers, like fqcodeldequeue. Let's make it idempotent to ease qdisctreereducebacklog callers' life: 1. updatevf decreases...

DEBIAN-CVE-2025-38177

In the Linux kernel, the following vulnerability has been resolved: schhfsc: make hfscqlennotify idempotent hfscqlennotify is not idempotent either and not friendly to its callers, like fqcodeldequeue. Let's make it idempotent to ease qdisctreereducebacklog callers' life: 1. updatevf decreases...

Drug cartel hacked cameras and phones to spy on FBI and identify witnesses

The "El Chapo" Mexican drug cartel snooped on FBI personnel through hacked cameras, and listened in on their phone calls to identify and kill potential witnesses, the US Department of Justice has said. And seven years on, the Bureau's defenses against this kind of surveillance are still inadequat...

Holographic Projection and Cyber Attack Surface: a Physical Analogy for Digital Security

This article presents an in-depth exploration of the analogy between the Holographic Principle in theoretical physics and cyber attack surfaces in digital security. Building on concepts such as black hole entropy and AdS/CFT duality, it highlights how complex infrastructures project their...

PT-2025-37217

Name of the Vulnerable Software and Affected Versions: Linux kernel affected versions not specified Description: A race condition exists between disabling quotas and running the quota rescan ioctl in the btrfs subsystem. This can lead to a use-after-free of qgroup records from the fs info-qgroup...

Boosting Generative Adversarial Transferability with Self-Supervised Vision Transformer Features

The ability of deep neural networks DNNs come from extracting and interpreting features from the data provided. By exploiting intermediate features in DNNs instead of relying on hard labels, we craft adversarial perturbation that generalize more effectively, boosting black-box transferability...

On the Feasibility of Poisoning Text-To-Image AI Models Via Adversarial Mislabeling

Today's text-to-image generative models are trained on millions of images sourced from the Internet, each paired with a detailed caption produced by Vision-Language Models VLMs. This part of the training pipeline is critical for supplying the models with large volumes of high-quality image-captio...

Vulnerability Disclosure through Adaptive Black-Box Adversarial Attacks on NIDS

Adversarial attacks, wherein slight inputs are carefully crafted to mislead intelligent models, have attracted increasing attention. However, a critical gap persists between theoretical advancements and practical application, particularly in structured data like network traffic, where...

Assessing Risk of Stealing Proprietary Models for Medical Imaging Tasks

The success of deep learning in medical imaging applications has led several companies to deploy proprietary models in diagnostic workflows, offering monetized services. Even though model weights are hidden to protect the intellectual property of the service provider, these models are exposed to...

Telegram Purged Chinese Crypto Scam Markets—Then Watched as They Rebuilt

Last month, Telegram banned black markets that sold tens of billions of dollars in crypto scam-related services. Now, as those markets rebrand and bounce back, it’s done nothing to stop them...

Pushing the Limits of Safety: a Technical Report on the ATLAS Challenge 2025

Multimodal Large Language Models MLLMs have enabled transformative advancements across diverse applications but remain susceptible to safety threats, especially jailbreak attacks that induce harmful outputs. To systematically evaluate and improve their safety, we organized the Adversarial Testing...

Alphabet Index Mapping: Jailbreaking LLMs through Semantic Dissimilarity

Large Language Models LLMs have demonstrated remarkable capabilities, yet their susceptibility to adversarial attacks, particularly jailbreaking, poses significant safety and ethical concerns. While numerous jailbreak methods exist, many suffer from computational expense, high token usage, or...