2026-08-10 归档

今日共收录 768 条安全情报,包含 81 个 CVE,374 篇安全通告,以及 66 篇研究论文。

← 返回归档列表
学术论文2026-08-06 发布INFO

Reversible Unlearnable Examples: Towards the Copyright Protection in Deep Learning Era

本文针对深度学习时代数据版权保护面临的两大威胁:未经授权的模型训练(非法利用数据训练模型)和恶意数据泄露(攻击者窃取或泄露受保护数据)。现有方法通常分别使用不可学习样本(unlearnable examples)和数字水印技术,但作者发现简

历史摘要 · 未重新核验 · arXiv cs.CR

学术论文2026-08-06 发布INFO

dfence: Fine-Grained Speculation Barriers for Efficient and Effective Hardware-Software Protection in the Spectre Era (Extended Version)

本文提出了一种名为 dfence 的新型 CPU 指令,旨在以硬件-软件协同的方式高效缓解推测执行攻击(如 Spectre-PHT 和 Spectre-STL)。现有的软件缓解方案(如 Speculative Load Hardening,

历史摘要 · 未重新核验 · arXiv cs.CR

学术论文2026-08-06 发布INFO

Tool Demo: Topology analysis with GPML for detection of cyberattacks in Water Distribution Networks

本文针对水分配网络(WDN)中工业控制系统(ICS)面临的网络攻击问题,提出一种基于拓扑分析的攻击检测方法。研究背景是传统检测方法依赖原始流量或协议信息,难以捕捉攻击引起的结构性变化,导致检测能力有限。作者采用基于图处理机器学习(GPML)

历史摘要 · 未重新核验 · arXiv cs.CR

学术论文2026-08-06 发布INFO

An End-to-End Threat Model for the Quantum-as-a-Service Pipeline

该论文针对量子计算即服务(Quantum-as-a-Service, QaaS)平台(如 IBM Quantum、IonQ Cloud、Amazon Braket)的端到端威胁建模问题展开研究。随着云化量子计算日益普及,混合量子-经典算法(

历史摘要 · 未重新核验 · arXiv cs.CR

学术论文2026-08-06 发布INFO

On the Figures of Merit for Quantum Software Security: Toward a Benchmarking Rubric

本文针对量子软件安全度量缺失的问题,提出了“量子软件安全品质因数(Security Figures of Merit, S-FoMs)”的初步框架。研究背景是:量子软件日益通过多租户和云端的量子即服务(QaaS)栈交付,近年研究已展示出跨流

历史摘要 · 未重新核验 · arXiv cs.CR

学术论文2026-08-06 发布INFO

Quantum One-Way Functions and Related Cryptographic Primitives

本文是一篇关于量子单向函数及相关量子态原语的综述性论文,由Georgios M. Nikolopoulos撰写,属于量子物理与数学物理交叉领域。论文指出,在经典密码学中,单向函数几乎支撑了所有标准密码协议,但量子密码学中除密钥分发外的原语研

历史摘要 · 未重新核验 · arXiv cs.CR

学术论文2026-08-06 发布INFO

Algebraic Cryptanalytic Extraction on Hard-Label Neural Networks

本文针对神经网络模型提取攻击在硬标签(hard-label)场景下的计算效率瓶颈,提出了一种基于代数方法的近似签名向量(ASV)技术。现有最优攻击由 Carlini 等人在 EUROCRYPT 2025 提出,理论上具有多项式时间复杂度,但

历史摘要 · 未重新核验 · arXiv cs.CR

学术论文2026-08-06 发布INFO

Detecting Safety Training Modification in Language Models via Activation Analysis

本文提出了一种名为 AMS(Activation-based Model Scanner,基于激活的模型扫描器)的工具,用于检测语言模型中的安全训练修改。其核心思想是:安全训练会在激活空间中为有害内容与良性内容的概念形成可测量的几何结构分离

历史摘要 · 未重新核验 · arXiv cs.CR

学术论文2026-08-06 发布INFO

A Note on the Influence of a Zero Length Nonce on GCM and GMAC

本文针对 GCM 和 GMAC 认证加密模式提出一种基于零长度 nonce 的简单攻击方法。作者指出,在 ISO/IEC 标准版本的 GCM/GMAC 中,nonce 被允许为空字符串(零长度),而 NIST 版本则明确要求 nonce 至

历史摘要 · 未重新核验 · arXiv cs.CR

漏洞预警2026-08-10 发布MEDIUM

A vulnerability was found in code-projects Task Management System 1.0. This issue affects some unknown processing of the file /user/CommentSave.php. The manipulation of the argu...

CVE-2026-19378

CVE-2026-19378 是 code-projects 出品的 Task Management System(任务管理系统)1.0 版本中存在的一处跨站脚本(XSS)漏洞。该漏洞位于 /user/CommentSave.php 文件,

历史摘要 · 未重新核验 · NVD

漏洞预警2026-08-09 发布HIGH

Improper Neutralization of Special Elements in Data Query Logic vulnerability in ash-project ash allows an attacker to forge a relationship to a record they cannot name, and to ...

CVE-2026-70395

CVE-2026-70395 是 ash-project ash 框架中的一个数据查询逻辑特殊元素未中和漏洞。当开发者对 belongs_to 关系使用 manage_relationship 并配置 on_lookup: :relate

历史摘要 · 未重新核验 · NVD

漏洞预警2026-08-09 发布INFO

Uncontrolled Resource Consumption vulnerability in ash-project ash allows an attacker to exhaust the memory of the node via a crafted keyset pagination cursor. Read actions wit...

CVE-2026-69659

Uncontrolled Resource Consumption vulnerability in ash-project ash allows an attacker to exhaust the memory of the node

历史摘要 · 未重新核验 · NVD

漏洞预警2026-08-09 发布INFO

A vulnerability was determined in itsourcecode Hospital Management System 1.0. The affected element is an unknown function of the file /viewdoctorconsultancycharge.php. This man...

CVE-2026-19364

A vulnerability was determined in itsourcecode Hospital Management System 1.0. The affected element is an unknown functi

历史摘要 · 未重新核验 · NVD

漏洞预警2026-08-09 发布INFO

A vulnerability was found in lmammino oidc-authorizer up to 0.4.0. Impacted is the function unwrap of the file src/handler.rs of the component Fixed Message Handler. The manipul...

CVE-2026-19363

A vulnerability was found in lmammino oidc-authorizer up to 0.4.0. Impacted is the function unwrap of the file src/handl

历史摘要 · 未重新核验 · NVD

漏洞预警2026-08-09 发布INFO

A vulnerability has been found in lmammino oidc-authorizer 0.4.0. This issue affects the function parse_token_from_header of the file src/parse_token_from_header.rs of the compo...

CVE-2026-19362

A vulnerability has been found in lmammino oidc-authorizer 0.4.0. This issue affects the function parse_token_from_heade

历史摘要 · 未重新核验 · NVD

漏洞预警2026-08-09 发布INFO

A flaw has been found in macrozheng mall 0504e86. This vulnerability affects unknown code of the file /sso/getAuthCode of the component mall-portal Module. Executing a manipulat...

CVE-2026-19361

A flaw has been found in macrozheng mall 0504e86. This vulnerability affects unknown code of the file /sso/getAuthCode o

历史摘要 · 未重新核验 · NVD

漏洞预警2026-08-09 发布INFO

Perl versions through 5.45.1 have out-of-bounds heap reads and writes during regular expression matching via an undersized superlinear cache in S_regmatch. The regex engine's s...

CVE-2026-15534

Perl versions through 5.45.1 have out-of-bounds heap reads and writes during regular expression matching via an undersiz

历史摘要 · 未重新核验 · NVD

漏洞预警2026-08-09 发布INFO

A vulnerability was detected in wongcyrus ExcelLexBot up to 0.0.3. This affects the function ExcelLexBotS3TriggerFunction of the component Lambda Function Handler. Performing a ...

CVE-2026-19360

A vulnerability was detected in wongcyrus ExcelLexBot up to 0.0.3. This affects the function ExcelLexBotS3TriggerFunctio

历史摘要 · 未重新核验 · NVD

漏洞预警2026-08-09 发布INFO

A security vulnerability has been detected in nxp-auto-goldvip gvip up to 1.4.0. Affected by this issue is the function SitewiseCustomFunction of the component Lambda Function H...

CVE-2026-19359

A security vulnerability has been detected in nxp-auto-goldvip gvip up to 1.4.0. Affected by this issue is the function

历史摘要 · 未重新核验 · NVD

漏洞预警2026-08-09 发布INFO

A weakness has been identified in 3CORESec Trapdoor up to 1.2.2. Affected by this vulnerability is the function DefaultFunction. This manipulation causes improper access control...

CVE-2026-19358

A weakness has been identified in 3CORESec Trapdoor up to 1.2.2. Affected by this vulnerability is the function DefaultF

历史摘要 · 未重新核验 · NVD