Aggregator

A new report from Bitdefender reveals the Russian-linked hacking group Curly COMrades is targeting Eastern Europe with a…

An unprecedented surge in brute-force attacks targeting Fortinet SSL VPN infrastructure, with over 780 unique IP addresses participating in coordinated assault campaigns.  The August 3rd attack represents the highest single-day volume recorded on GreyNoise’s Fortinet SSL VPN Bruteforcer tag in recent months, raising concerns about potential zero-day vulnerabilities and sophisticated threat actor operations. Key Takeaways1. […]

The post Hackers Attacking Fortinet SSL VPN Under Attack From 780 unique IPs appeared first on Cyber Security News.

Assisted Living Pharmacy Service LLC Falls Victim to Qilin Ransomware

IBM’s 2025 Cost of a Data Breach Report drives home that point: attackers today are often “logging in rather than hacking in".

The post IBM’s Cost of a Data Breach Report 2025 appeared first on Security Boulevard.

前美国国家安全局（NSA）局长、中央安全局局长和网战司令部司令、现 OpenAI 董事、退役陆军上将保罗·仲宗根（Paul Nakasone）在拉斯维加斯出席 Defcon 网络安全大会，与 Defcon 创始人 Jeff Moss 谈论了美国当前的政治气候。特朗普当局解雇了被认为不忠诚的网络安全官员，撤销了前 CISA 主任 Chris Krebs 和 Jen Easterly 的安全许可，仲宗根评论说，美国科技公司在 2025 和 2026 年试图保持中立将会非常非常困难。

Researchers just proved LLMs can autonomously plan and execute full-scale cyberattacks — turning AI from a tool into an active threat actor.

The post And Now, LLMs Don’t Need Human Intervention to Plan and Execute Large, Complex Attacks  appeared first on Security Boulevard.

Ivanti has released critical security updates addressing multiple high and medium-severity vulnerabilities across its Connect Secure, Policy Secure, and Zero Trust Access (ZTA) gateway products.  The vulnerabilities, identified through internal discovery and responsible disclosure programs, could enable remote attackers to trigger denial-of-service (DoS) attacks without authentication, though no active exploitation has been detected at the […]

The post Ivanti Connect Secure, Policy Secure and ZTA Vulnerabilities Let Attackers Trigger DoS Attack appeared first on Cyber Security News.

Cybersecurity researchers at GreyNoise have detected an alarming surge in brute-force attacks against Fortinet SSL VPN systems, with over 780 unique IP addresses launching coordinated attacks in a single day—marking the highest daily volume recorded for this type of attack in recent months. The sophisticated campaign appears to represent a significant escalation in targeting Fortinet’s […]

The post Fortinet SSL VPN Targeted by Hackers from 780 Unique IP Addresses appeared first on GBHackers Security | #1 Globally Trusted Cyber Security News Platform.

ЛДПР предложила блокировать мат в соцсетях и на сайтах.

Over 29,000 Microsoft Exchange servers remain unpatched against a vulnerability that could allow attackers to seize control of entire domains in hybrid cloud environments

Creator, Author and Presenter: Wendy Nather

Our deep appreciation to Security BSides - San Francisco and the Creators, Authors and Presenters for publishing their BSidesSF 2025 video content on YouTube. Originating from the conference’s events held at the lauded CityView / AMC Metreon - certainly a venue like no other; and via the organization's YouTube channel.

Additionally, the organization is welcoming volunteers for the BSidesSF Volunteer Force, as well as their Program Team & Operations roles. See their succinct BSidesSF 'Work With Us' page, in which, the appropriate information is to be had!

Permalink

The post BSidesSF 2025: Fire, Brimstone And Bad Security Decisions appeared first on Security Boulevard.

Ivanti has released critical security updates addressing multiple vulnerabilities in its Connect Secure, Policy Secure, and ZTA Gateway products that could allow remote attackers to launch denial-of-service attacks. The company disclosed four vulnerabilities on August 12, 2025, with CVSS scores ranging from medium to high severity, though no active exploitation has been detected at the […]

The post Ivanti Connect Secure, Policy Secure, and ZTA Flaws Allow Attackers to Launch DoS Attacks appeared first on GBHackers Security | #1 Globally Trusted Cyber Security News Platform.

Jenkins disclosed CVE-2025-53652, also known as SECURITY-3419, as part of a batch of 31 plugin vulnerabilities. Initially rated as medium severity, this flaw affects the Git Parameter plugin and was described merely as enabling attackers to inject arbitrary values into Git parameters a description that understated its potential for severe exploitation. However, deeper analysis reveals […]

The post 1,500 Jenkins Servers Vulnerable to Command Injection via Git Parameter Plugin appeared first on GBHackers Security | #1 Globally Trusted Cyber Security News Platform.

本文链接

腾讯 TEG 安全平台部代码安全负责人张栋已确认出席，并发表题为《大模型驱动安全升级：腾讯代码安全应用实践》的主题演讲。安全漏洞是每一个企业都必须面对的问题，处理不好动辄影响服务可用性、导致核心业务敏感数据泄露，对业务及公司造成不良影响。基于此在企业研发体系中，绝大多数企业会选择通过安全左移在软件开发生命周期的早期阶段进行风险识别和安全问题修复。但是，传统的静态分析工具本身因其检测方式导致的覆盖有限、高误报、修复效率低等问题，增大了业务团队负担，致使业务对安全问题响应慢、体感差。在本次分享中，张栋将结合腾讯内部的实践经验，深入探讨大模型在代码安全场景中的应用，分享在不同发展阶段其遇到的挑战及其解决方案。

演讲提纲 1.代码安全现状及问题分析

• 传统代码安全现状
• 传统代码安全痛点
• 业务安全场景痛点
• 传统代码安全的局限性

2. 大模型在代码安全场景应用与挑战

• 大模型赋能传统工具
• 大模型助力安全左移
• 大模型自动化漏洞检测

3. 技术架构及解决方案

• 行业前沿解析
• 架构设计及能力演进
• 基础能力建设

4. 总结及展望

• 工程实践心得
• 大模型在代码安全场景落地的方向

演讲亮点

• 代码安全场景大模型落地思路及过程经验分享
• 大模型自动化漏洞扫描思路及过程经验分享
• 专用领域大模型的微调思路及过程经验分享

听众收益

• 了解专用模型领域微调思路及落地实践
• 了解大模型落地过程中常见问题及解决方案
• 了解大模型赋能安全的基础上，进一步泛化至其他应用场景

本文链接

PyTorch 是一个基于 Torch 库的机器学习库，广泛应用于计算机视觉和自然语言处理等领域。它是最受欢迎的深度学习框架之一。

然而，尽管 PyTorch 拥有强大的功能，但其背后潜藏着潜在的安全风险。最初，PyTorch 使用 pickle 保存模型，但由于 pickle 反序列化的不安全性，加载模型时存在远程代码执行（RCE）的风险。随后，PyTorch 引入了 weights_only 参数来增强安全性。官方文档指出，weights_only=True 被认为是安全的，并建议使用它，而不是 weights_only=False。

多年来，weights_only=True 的安全性一直没有受到挑战。然而，我们的研究揭示了一些令人不安的事实。我们发现，使用 weights_only=True 的 torch.load 支持 TorchScript，这促使我们深入研究 TorchScript 的内部工作机制。在一段时间的研究后，我们发现了几个漏洞，并最终实现了远程代码执行（RCE）。我们迅速将这一发现报告给了 PyTorch，PyTorch 确认了该漏洞并为我们分配了 CVE-2025-32434。这个发现颠覆了已有的认知，并对许多 AI 应用产生了深远的影响。我们将在本报告中深入分析该漏洞的影响。

在本次简报中，我们将介绍如何获得灵感并发现这一有趣的漏洞。同时，我们的发现再次验证了这句话：“你曾以为的安全港，实际上是敌人的水域。”

The first documented deployment of the novel malware in a campaign against the Middle Eastern public sector and aviation industry may be tied to China's state-sponsored actor Earth Baxia.

A new cyber-espionage threat group has been using a new backdoor malware that provides persistent access through a seemingly inactive scheduled task. [...]

FortiGuard Labs has reported a dramatic spike in exploitation attempts targeting CitrixBleed 2, a critical buffer over‑read flaw (CVE‑2025‑5777) affecting Citrix NetScaler ADC (Application Delivery Controller) and Gateway devices. Since July 28, 2025, they have detected over 6,000 exploitation attempts, mostly in the US, Australia, Germany and the UK, “with adversaries primarily focusing on high-value sectors such as technology, banking, healthcare, and education.” Meanwhile, the Dutch National Cyber Security Centre (NCSC‑NL) has confirmed that another NetScaler … More →

The post Netscaler vulnerability was exploited as zero-day for nearly two months (CVE-2025-6543) appeared first on Help Net Security.

有 133 年历史的柯达公司警告投资者该公司可能停止运营。柯达在最新财报中称，它缺乏资金去偿还约 5 亿美元的即将到期的债务。柯达计划通过停止支付退休金计划筹集现金，该公司的产品如相机、墨水和胶卷都在美国生产，因此关税不会对其业务产生影响。柯达（Eastman Kodak）成立于 1892 年，创始人 George Eastman 在 1888 年出售了第一台柯达相机。柯达的名字是 Eastman 生造出来的。柯达一度占据了美国 90% 的胶卷和 85% 的相机销量，虽然它研制出了第一台数码相机，但却将其拱手让出，最终在数字摄影时代落后，2012 年申请破产保护，2020 年转型为制药原料生产商获得了短暂的喘息。

Over 3,300 Citrix NetScaler devices remain unpatched against a critical vulnerability that allows attackers to bypass authentication by hijacking user sessions, nearly two months after patches were released. [...]