PT-2025-32167 · Nvidia · Nvidia Triton Inference Server

Name of the Vulnerable Software and Affected Versions: NVIDIA Triton Inference Server for Windows and Linux affected versions not specified Description: NVIDIA Triton Inference Server contains an issue where a user can trigger a segmentation fault by submitting an invalid request that causes...

PT-2025-32159 · Nvidia · Nvidia Triton Inference Server

Name of the Vulnerable Software and Affected Versions: NVIDIA Triton Inference Server affected versions not specified Description: NVIDIA Triton Inference Server contains an issue in the Python backend that could allow an attacker to cause an out-of-bounds write. Successful exploitation of this...

PT-2025-31924

Name of the Vulnerable Software and Affected Versions NVIDIA Triton Inference Server versions prior to 25.07 Description NVIDIA Triton Inference Server contains a flaw where an attacker can trigger a stack overflow through crafted HTTP requests. Successful exploitation could lead to remote code...

PT-2025-32164 · Nvidia · Nvidia Triton Inference Server

Name of the Vulnerable Software and Affected Versions: NVIDIA Triton Inference Server for Windows and Linux affected versions not specified Description: The NVIDIA Triton Inference Server contains a flaw that allows an attacker to trigger uncontrolled recursion with a crafted input. Successful...

PT-2025-32166 · Nvidia · Nvidia Triton Inference Server

Name of the Vulnerable Software and Affected Versions: NVIDIA Triton Inference Server for Windows and Linux affected versions not specified Description: NVIDIA Triton Inference Server is susceptible to an integer overflow triggered by specially crafted inputs. Successful exploitation of this issu...

PT-2025-32160 · Nvidia · Nvidia Triton Inference Server

Name of the Vulnerable Software and Affected Versions: NVIDIA Triton Inference Server for Windows and Linux affected versions not specified Description: The NVIDIA Triton Inference Server contains a flaw that allows a user to trigger a divide-by-zero error by submitting a malformed request...

PT-2025-32161 · Nvidia · Nvidia Triton Inference Server

Name of the Vulnerable Software and Affected Versions: NVIDIA Triton Inference Server affected versions not specified Description: NVIDIA Triton Inference Server contains an issue where multiple requests can cause a double free when a stream is cancelled before processing. Successful exploitation...

Hot-Swap MarkBoard: an Efficient Black-Box Watermarking Approach for Large-Scale Model Distribution

Recently, Deep Learning DL models have been increasingly deployed on end-user devices as On-Device AI, offering improved efficiency and privacy. However, this deployment trend poses more serious Intellectual Property IP risks, as models are distributed on numerous local devices, making them...

Program Analysis for High-Value Smart Contract Vulnerabilities: Techniques and Insights

A widespread belief in the blockchain security community is that automated techniques are only good for detecting shallow bugs, typically of small value. In this paper, we present the techniques and insights that have led us to repeatable success in automatically discovering high-value smart...

Cascading and Proxy Membership Inference Attacks

A Membership Inference Attack MIA assesses how much a trained machine learning model reveals about its training data by determining whether specific query instances were included in the dataset. We classify existing MIAs into adaptive or non-adaptive, depending on whether the adversary is allowed...

LoRA-Leak: Membership Inference Attacks against LoRA Fine-Tuned Language Models

Language Models LMs typically adhere to a "pre-training and fine-tuning" paradigm, where a universal pre-trained model can be fine-tuned to cater to various specialized domains. Low-Rank Adaptation LoRA has gained the most widespread use in LM fine-tuning due to its lightweight computational cost...

Tab-MIA: a Benchmark Dataset for Membership Inference Attacks on Tabular Data in LLMs

Large language models LLMs are increasingly trained on tabular data, which, unlike unstructured text, often contains personally identifiable information PII in a highly structured and explicit format. As a result, privacy risks arise, since sensitive records can be inadvertently retained by the...

Learning-Based Privacy-Preserving Graph Publishing against Sensitive Link Inference Attacks

Publishing graph data is widely desired to enable a variety of structural analyses and downstream tasks. However, it also potentially poses severe privacy leakage, as attackers may leverage the released graph data to launch attacks and precisely infer private information such as the existence of...

CompLeak: Deep Learning Model Compression Exacerbates Privacy Leakage

Model compression is crucial for minimizing memory storage and accelerating inference in deep learning DL models, including recent foundation models like large language models LLMs. Users can access different compressed model versions according to their resources and budget. However, while existi...

When LLMs Copy to Think: Uncovering Copy-Guided Attacks in Reasoning LLMs

Large Language Models LLMs have become integral to automated code analysis, enabling tasks such as vulnerability detection and code comprehension. However, their integration introduces novel attack surfaces. In this paper, we identify and investigate a new class of prompt-based attacks, termed...

GATEBLEED: Exploiting On-Core Accelerator Power Gating for High Performance and Stealthy Attacks on AI

As power consumption from AI training and inference continues to increase, AI accelerators are being integrated directly into the CPU. Intel's Advanced Matrix Extensions AMX is one such example, debuting on the 4th generation Intel Xeon Scalable CPU. We discover a timing side and covert channel,...

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...

HyDRA: a Hybrid Dual-Mode Network for Closed- and Open-Set RFFI with Optimized VMD

Device recognition is vital for security in wireless communication systems, particularly for applications like access control. Radio Frequency Fingerprint Identification RFFI offers a non-cryptographic solution by exploiting hardware-induced signal distortions. This paper proposes HyDRA, a Hybrid...

FacialMotionID: Identifying Users of Mixed Reality Headsets Using Abstract Facial Motion Representations

Facial motion capture in mixed reality headsets enables real-time avatar animation, allowing users to convey non-verbal cues during virtual interactions. However, as facial motion data constitutes a behavioral biometric, its use raises novel privacy concerns. With mixed reality systems becoming...

LRCTI: a Large Language Model-Based Framework for Multi-Step Evidence Retrieval and Reasoning in Cyber Threat Intelligence Credibility Verification

Verifying the credibility of Cyber Threat Intelligence CTI is essential for reliable cybersecurity defense. However, traditional approaches typically treat this task as a static classification problem, relying on handcrafted features or isolated deep learning models. These methods often lack the...