Хакерам даже не нужно стараться. Названы три причины, по которым взломщики до сих пор сильнее корпораций
Самые тревожные выводы из нового масштабного исследования Hunt & Hackett.
Aggregator
Metacritic 不收录 AI 生成的评测
3 months 3 weeks ago
Metacritic 是一家聚合电影、电视、音乐专辑、游戏评测及其评分的网站，它会基于相关评分给出一个加权平均值。Metacritic 成立于 2001 年，至今有 25 年历史，它的综合评分获得了广泛认可。网站联合创始人 Marc Doyle 在一份声明中表示 Metacritic 不会收录 AI 生成的评测。在这之前，一家叫 Videogamer.com 的英国老牌游戏网站（曾经很受欢迎）在被博彩公司收购之后解雇了大部分员工，然后用 AI 生成了新游戏的评测和评分。该网站评测作者的肖像是用 ChatGPT 生成的。Metacritic 在获悉之后删除了该网站的评测。
NightSpire
3 months 3 weeks ago
You must login to view this content
cohenido
NightSpire
3 months 3 weeks ago
You must login to view this content
cohenido
OpenAI Will Deploy AI in US Military Classified Networks
3 months 3 weeks ago
Announcement Comes Hours After Trump Blacklists Anthropic
OpenAI said late Friday night it reached an agreement with the U.S. Department of Defense to deploy its large language models onto military classified networks. The announcement came hours after President Donald Trump instructed federal agencies to cease using AI developed by OpenAI rival Anthropic.
OpenAI said late Friday night it reached an agreement with the U.S. Department of Defense to deploy its large language models onto military classified networks. The announcement came hours after President Donald Trump instructed federal agencies to cease using AI developed by OpenAI rival Anthropic.
US and Israel Launch 'Major Combat Operations' Against Iran
3 months 3 weeks ago
Trump Says Combat Targets Regime in Tehran, As Iran Responds With Missile Attacks
U.S. President Donald Trump announced the launch of "major combat operations in Iran," in coordination with Israel, targeting the regime in Tehran over its nuclear ambitions. Iran responded with missile attacks, as cybersecurity experts forecast online reprisals.
U.S. President Donald Trump announced the launch of "major combat operations in Iran," in coordination with Israel, targeting the regime in Tehran over its nuclear ambitions. Iran responded with missile attacks, as cybersecurity experts forecast online reprisals.
Building an AI Agent for Adaptive MFA Decisioning
3 months 3 weeks ago
Build an AI agent for adaptive MFA decisioning using risk-based authentication, machine learning, and intelligent security automation.
The post Building an AI Agent for Adaptive MFA Decisioning appeared first on Security Boulevard.
MojoAuth - Advanced Authentication & Identity Solutions
Как заставить пропагандиста признаться, что он робот. Инструкция от пользователей Threads
3 months 3 weeks ago
Простой способ проверить, с кем вы спорите в Интернете.
CVE-2026-1542 | Super Stage WP Plugin up to 1.0.1 on WordPress Gadget deserialization (EUVD-2026-9099)
3 months 3 weeks ago
A vulnerability identified as critical has been detected in Super Stage WP Plugin up to 1.0.1 on WordPress. Affected is an unknown function of the component Gadget Handler. This manipulation causes deserialization.
The identification of this vulnerability is CVE-2026-1542. It is possible to initiate the attack remotely. There is no exploit available.
vuldb.com
CVE-2026-24714 | Netgear PR2000 TelnetEnable inclusion of undocumented features or chicken bits (EUVD-2026-9100)
3 months 3 weeks ago
A vulnerability identified as problematic has been detected in Netgear PR2000. The impacted element is an unknown function of the component TelnetEnable. This manipulation causes inclusion of undocumented features or chicken bits. This vulnerability only affects products that are no longer supported by the maintainer.
This vulnerability is handled as CVE-2026-24714. The attack can be initiated remotely. There is not any exploit available.
vuldb.com
CREST CTF
3 months 3 weeks ago
Name: CREST CTF (an CREST CTF event.)
Date: Feb. 28, 2026, 4:30 a.m. — 28 Feb. 2026, 12:30 UTC [add to calendar]
Format: Jeopardy
On-site
Location: India, Pune
Offical URL: https://crestpcu.in/
Rating weight: 6.00
Event organizers: CREST
Date: Feb. 28, 2026, 4:30 a.m. — 28 Feb. 2026, 12:30 UTC [add to calendar]
Format: Jeopardy
On-site
Location: India, Pune
Offical URL: https://crestpcu.in/
Rating weight: 6.00
Event organizers: CREST
NASA 宣布 Artemis III 任务仍然是绕月而非登月
3 months 3 weeks ago
NASA 宣布调整阿耳忒弥斯(Artemis)系列载人登月计划：原计划今年 3 月 6 日发射的 Artemis II 载人绕月飞行任务推迟到最早 4 月 1 日进行；Artemis III 载人登月任务不再计划登月，而仍然是绕月飞行。NASA 局长 Jared Isaacman 表示，更为循序渐进的新计划让 NASA 团队能测试飞行和改进技术。
2025 年度总结
3 months 3 weeks ago
老伙计们，老鸽子来也！
Tr0y
External Authentication: Exploring WS-Trust for Authentication
3 months 3 weeks ago
Learn how WS-Trust powers external authentication in hybrid identity environments. Explore the Security Token Service (STS) and legacy bridge protocols.
The post External Authentication: Exploring WS-Trust for Authentication appeared first on Security Boulevard.
SSOJet - Enterprise SSO & Identity Solutions
5 IoT Vulnerabilities That Stop Projects and How to Avoid Them
3 months 3 weeks ago
Stop the 75% failure rate. Learn which device vulnerabilities stall deployments and the exact fixes that get IoT projects to production.
Owais Sultan
AI网络爬虫安全白皮书
3 months 3 weeks ago
Author: Guancheng Li and Zheng Wang of Tencent Xuanwu Lab
本文是腾讯玄武实验室发布的《AI网络爬虫安全白皮书》。我们系统分析了 AI 时代服务端浏览器 / 爬虫在真实业务中的典型使用方式，以及由此带来的新的攻击面与风险。
在这篇白皮书中，我们结合多款实际产品的安全测试案例，梳理了完整的攻击链路，并提出以“静态攻击面收敛 + 动态行为隔离”为核心的纵深防御框架，同时给出了开源防护方案 SEChrome，帮助企业在工程上落地安全的服务端浏览器架构。
下载链接：AI网络爬虫安全白皮书.pdf
xlab
ComfyUI-Manager 远程代码执行风险通告
3 months 3 weeks ago
近期腾讯玄武实验室发现可视化 AI 工作流工具 ComfyUI 的官方扩展组件 ComfyUI-Manager 中存在一个高危漏洞（CVE-2025-67303）。利用该漏洞可在无需任何账号的情况下远程入侵安装 ComfyUI 的系统。玄武实验室在发现漏洞后向 ComfyUI 官方进行了报告，目前该漏洞已被修复。
xlab
量子计算机距离攻破 RSA-2048 还有多远
3 months 3 weeks ago
Author: Guancheng Li of Tencent Xuanwu Lab
在当今数字世界中，RSA‑2048 与 ECC 等经典公钥密码是最广泛应用的加密标准，支撑着网络安全、金融交易和隐私保护的底层信任。然而，这一基石正面临量子计算的潜在威胁。理论上，量子计算机能够以远快于经典计算机的速度分解大整数和离散对数求解，从而在短时间内破解 RSA 和 ECC 加密。这一前景既令人兴奋，也令人担忧。
问题在于：量子计算机的发展究竟到了什么阶段？有人乐观地认为经典公钥密码的“倒计时”已经开始；也有人怀疑，受限于制造难度，真正可用的量子计算机还遥遥无期。市面上相关论调不一，往往乐观或悲观，但核心疑问始终萦绕：量子计算机距离破解经典公钥密码还有多远？
xlab
我们的AI发现了一个零知识证明库的漏洞，Sam Altman的项目也用了这个库
3 months 3 weeks ago
Author: Guancheng Li, Xiaolin Zhang and Yang Yu of Tencent Xuanwu Lab
2025 年 8 月，腾讯玄武实验室的阿图因自动化漏洞挖掘引擎在零知识证明库 gnark 中发现了一个高危漏洞（CVE-2025-57801，CVSS 8.6）。之后，玄武实验室联合上海交通大学 GOSSIP 实验室及郁昱教授团队共同完成了漏洞复现。
xlab
玄武在BlackHat揭示劫持智能体达成RCE的新方法
3 months 3 weeks ago
Author: Jiashuo Liang and Guancheng Li of Tencent Xuanwu Lab
0x00 前言大语言模型（LLM）正在从简单的对话工具演化为能够编写代码、操作浏览器、执行系统命令的智能体。随着大模型应用的演进，提示词注入攻击的威胁也在不断升级。
设想这样一个场景：你让AI助手帮你编写代码，它却突然开始执行恶意指令，控制了你的电脑。这种看似科幻的情节，如今正在变为现实。
本文将介绍一种新型的提示词注入攻击范式。攻击者只需掌握一套“通用触发器”，就能精确控制大模型输出任意攻击者指定的内容，从而利用AI智能体实现远程代码执行等高风险操作。
xlab