Aggregator
谷歌在上海办了场 AI 嘉年华,开发者们却说像逛「AI 基地」
价值1.4 W人民币漏洞!骚!
Majority of Organizations Ship Vulnerable Code, Study Finds
CVE-2025-8800 | Open5GS up to 2.7.5 AMF src/mme/esm-handler.c esm_handle_pdn_connectivity_request denial of service (Issue 3980 / EUVD-2025-24080)
CVE-2025-7778 | Icons Factory Plugin up to 1.6.12 on WordPress delete_files authorization
CVE-2025-7650 | BizCalendar Web Plugin up to 1.1.0.50 on WordPress bizcalv file inclusion
CVE-2025-8905 | Inpersttion for Theme Plugin up to 1.0 on WordPress theme_section_shortcode code injection
CVE-2025-8091 | EventON Lite Plugin up to 2.4.6 on WordPress Shortcode add_eventon information disclosure
CVE-2025-7507 | elink Plugin up to 1.1.0 on WordPress input validation
CVE-2025-7662 | Gestion de tarifs Plugin up to 1.4 on WordPress Shortcode intitule sql injection
CVE-2025-8080 | Alobaidi Captcha Plugin up to 1.0.3 on WordPress Setting cross site scripting
CVE-2025-8720 | Plugin README Parser Plugin up to 1.3.15 on WordPress target cross site scripting
CVE-2025-5844 | Radius Blocks Plugin up to 2.2.1 on WordPress subHeadingTagName cross site scripting
New NIST Concept Paper Outlines AI-Specific Cybersecurity Framework
AI Threats & Adaptive Authentication: How to Be Protected Against Deepfakes and Credential Attacks
Learn how adaptive authentication defends against deepfakes, credential attacks, and AI threats to keep your business secure.
The post AI Threats & Adaptive Authentication: How to Be Protected Against Deepfakes and Credential Attacks appeared first on Security Boulevard.
【CVE-2025-27388】OPPO健康APP存在Webview任意URL加载导致Token泄露风险的致谢公告
【7心协力】日程过半!6家SRC活动仍在火热进行中!
VirtualBox 7.2 Adds Windows 11/Arm VM Support and Key Bug Fixes
Oracle has released VirtualBox 7.2.0, a major update that significantly expands ARM virtualization capabilities and introduces comprehensive Windows 11/ARM support. Released on August 14, 2025, this update represents a substantial leap forward in cross-platform virtualization technology, addressing long-standing limitations and introducing experimental features that could reshape how users approach virtual machine deployment across different architectures. […]
The post VirtualBox 7.2 Adds Windows 11/Arm VM Support and Key Bug Fixes appeared first on GBHackers Security | #1 Globally Trusted Cyber Security News Platform.
开源MTA搭建的邮件网关何谈安全?
HW 行动中,部分厂商因开源邮件网关组件漏洞沦陷,这一现象的背后,实则是开源 MTA 模块在技术根基上的诸多缺陷在实战中的集中暴露。
开源MTA的技术隐患:从架构缺陷到实战风险
国内邮件安全网关市场中,部分厂商因核心技术投入不足,产品能力分化明显,约70%邮件网关产品以 Postfix、Sendmail 等开源技术为根基,其功能扩展常走 “技术捷径”:实现反垃圾、病毒扫描等核心安全功能时,采用 “外部调用拼接式” 编程 —— 通过 Shell 脚本或命令行拼接调用外部安全引擎,而非原生 API 深度集成。这种开源改造存在三重硬伤:
1、效率损耗大:每封邮件处理需反复启动外部进程,高并发场景下 CPU 与内存占用激增,易成性能瓶颈。
2、安全攻击面扩大:命令行参数注入或转义不当,易被恶意邮件利用形成 Shell 注入风险。
3、稳定性和维护性差:进程间通信可靠性远低于原生库调用,错误链冗长难追踪。
根源在于研发模式不成熟。今年HW期间,多家机构网关产品即因开源 MTA 的内存管理缺陷、命令注入漏洞被攻破。更值得警惕的是,基于开源 MTA 搭建的网关,核心机制受制于人,难谈安全自主;且主流开源 MTA 多源自美国,不排除被植入后门的风险,一旦发生意外,可能导致信息泄露甚至系统失控。
CACTER自研破局:从底层重构邮件安全根基
CACTER邮件网关的破局关键,在于对 MTA核心组件实现 100% 自研,从协议处理到内存管理摆脱开源依赖,以此筑牢产品安全根基。
这一技术底气源自 26 年行业深耕经验与累计 63 项专利技术,始终以 “实战有效性”为导向—— 这正是CACTER在本次安全攻防中保持“零失陷”的核心秘诀。而除了这份自研硬实力外,CACTER的稳健表现,更离不开四大核心能力的坚实支撑:
26 年实战积累 从“被动防御” 到“主动适应”
邮件安全从不是 “闭门造车”,而是在借鉴与优化中动态应对实战挑战。26 年服务多行业的经验表明,稳定、适配、可迭代的方案,才是关键场景的坚实保障。
如今,AI 攻击推动邮件安全进入 “微秒级” 竞争时代。立足实战、融合优化的自研技术路线,已被证明是应对复杂威胁的可靠选择。26 年的经验积淀,硬核的产品实力,专业的服务团队,让CACTER在历次攻防中,交出让客户真正放心的答卷。