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...

DP2Guard: a Lightweight and Byzantine-Robust Privacy-Preserving Federated Learning Scheme for Industrial IoT

Privacy-Preserving Federated Learning PPFL has emerged as a secure distributed Machine Learning ML paradigm that aggregates locally trained gradients without exposing raw data. To defend against model poisoning threats, several robustness-enhanced PPFL schemes have been proposed by integrating...

DATABench: Evaluating Dataset Auditing in Deep Learning from an Adversarial Perspective

The widespread application of Deep Learning across diverse domains hinges critically on the quality and composition of training datasets. However, the common lack of disclosure regarding their usage raises significant privacy and copyright concerns. Dataset auditing techniques, which aim to...

UniAud: a Unified Auditing Framework for High Auditing Power and Utility with One Training Run

Differentially private DP optimization has been widely adopted as a standard approach to provide rigorous privacy guarantees for training datasets. DP auditing verifies whether a model trained with DP optimization satisfies its claimed privacy level by estimating empirical privacy lower bounds...

Generative AI for Vulnerability Detection in 6G Wireless Networks: Advances, Case Study, and Future Directions

The rapid advancement of 6G wireless networks, IoT, and edge computing has significantly expanded the cyberattack surface, necessitating more intelligent and adaptive vulnerability detection mechanisms. Traditional security methods, while foundational, struggle with zero-day exploits, adversarial...

Can One Safety Loop Guard Them All? Agentic Guard Rails for Federated Computing

We propose Guardian-FC, a novel two-layer framework for privacy preserving federated computing that unifies safety enforcement across diverse privacy preserving mechanisms, including cryptographic back-ends like fully homomorphic encryption FHE and multiparty computation MPC, as well as statistic...

CVE-2023-47297

A settings manipulation vulnerability in NCR Terminal Handler v1.5.1 allows attackers to execute arbitrary commands, including editing system security auditing configurations...

CVE-2023-47297

A settings manipulation vulnerability in NCR Terminal Handler v1.5.1 allows attackers to execute arbitrary commands, including editing system security auditing configurations...

CVE-2023-47297

A settings manipulation vulnerability in NCR Terminal Handler v1.5.1 allows attackers to execute arbitrary commands, including editing system security auditing configurations...

PT-2025-26608 · Ncr · Ncr Terminal Handler

Name of the Vulnerable Software and Affected Versions: NCR Terminal Handler version 1.5.1 Description: A settings manipulation issue allows attackers to execute arbitrary commands, including editing system security auditing configurations. Recommendations: For NCR Terminal Handler version 1.5.1,...

CVE-2023-47297

CVE-2023-47297: A settings manipulation vulnerability in NCR Terminal Handler v1.5.1 permits an attacker to execute arbitrary commands, including editing system security auditing configurations. CVSS v3.1 base score 9.8 (Network, Low complexity, No user interaction, Privileges=None; Impact: Confi...

Risks & Benefits of LLMs & GenAI for Platform Integrity, Healthcare Diagnostics, Cybersecurity, Privacy & AI Safety: a Comprehensive Survey, Roadmap & Implementation Blueprint

Large Language Models LLMs and generative AI GenAI systems such as ChatGPT, Claude, Gemini, LLaMA, and Copilot, developed by OpenAI, Anthropic, Google, Meta, and Microsoft are reshaping digital platforms and app ecosystems while introducing key challenges in cybersecurity, privacy, and platform...

Risks and Benefits of LLMs and GenAI for Platform Integrity, Healthcare Diagnostics, Cybersecurity, Privacy and AI Safety: a Comprehensive Survey, Roadmap and Implementation Blueprint

Large Language Models LLMs and generative AI GenAI systems such as ChatGPT, Claude, Gemini, LLaMA, and Copilot, developed by OpenAI, Anthropic, Google, Meta, and Microsoft are reshaping digital platforms and app ecosystems while introducing key challenges in cybersecurity, privacy, and platform...

Enhancing One-run Privacy Auditing with Quantile Regression-Based Membership Inference

Differential privacy DP auditing aims to provide empirical lower bounds on the privacy guarantees of DP mechanisms like DP-SGD. While some existing techniques require many training runs that are prohibitively costly, recent work introduces one-run auditing approaches that effectively audit DP-SGD...

CVE-2025-24288

The Versa Director software exposes a number of services by default and allow attackers an easy foothold due to default credentials and multiple accounts most with sudo access that utilize the same default credentials. By default, Versa director exposes ssh and postgres to the internet, alongside...

PT-2025-26196 · Versa · Versa Director

Name of the Vulnerable Software and Affected Versions: Versa Director affected versions not specified Description: The Versa Director software exposes several services by default, including ssh and postgres, which can provide an easy foothold for attackers due to default credentials and multiple...

The vulnerability of the Service Account Auditing service of the Windows Active Directory management and reporting software Zoho ManageEngine ADAudit Plus allows a perpetrator to execute arbitrary code.

The vulnerability of the Service Account Auditing service in the Windows Active Directory management and reporting software Zoho ManageEngine ADAudit Plus is related to the lack of security measures for the SQL query structure. Exploitation of this vulnerability could allow a malicious actor to...

CVE-2025-27709

Zohocorp ManageEngine ADAudit Plus versions 8510 and prior are vulnerable to authenticated SQL injection in the Service Account Auditing reports...

The vulnerability of the Service Account Auditing service of the Windows Active Directory management and reporting software Zoho ManageEngine ADAudit Plus allows a perpetrator to execute arbitrary code.

The vulnerability of the Service Account Auditing service in the Windows Active Directory management and reporting software Zoho ManageEngine ADAudit Plus is related to the lack of security measures for the SQL query structure. Exploitation of this vulnerability could allow a malicious actor to...

CVE-2025-49004

CVE-2025-49004 affects Caido prior to version 0.48.0. The issue arises from missing DNS rebinding protection, allowing a malicious site to load Caido on an attacker-controlled domain and hijack the authentication flow, potentially enabling remote command execution during the initial setup (and ev...