大语言模型幻觉驱动幻影域名抢注（Phantom Squatting）攻击机理与分层 ...

生成式大语言模型（LLM）普遍存在事实幻觉缺陷，模型会自主生成不存在、符合语义逻辑的域名地址。攻击者利用该稳定幻觉特征，提前注册此类虚构域名并部署仿冒站点实施钓鱼、恶意软件分发，该新型攻击被 Palo Alto Networks Unit 42 ...

供应链 CRM 数据泄露引发定向钓鱼多维检测与闭环防御研究

现有工程与学术研究存在两处明显短板：第一，多数供应链安全研究聚焦 OAuth 令牌防护、第三方供应商准入审计，未结合 CRM 数据泄露后的定向钓鱼衍生威胁开展联动分析；第二，现有钓鱼检测模型未引入客户泄露业务画像特征，无法区分普通泛化钓鱼与高欺骗性定向钓鱼，缺少适配 SaaS 企业客户防护的轻量化检测代码与落地流程。基于上述现实痛点，本文以 LastPass 供应链泄露事件为核心案例，系统拆解定向 ...
BBC

Anthropic accuses Chinese rival Alibaba of illicitly extracting AI capabilities

US artificial intelligence (AI) giant Anthropic has accused Chinese e-commerce and technology firm Alibaba of "brazenly" and "illicitly" extracting its Claude AI model's capabilities. In a letter sent ...
note

AI is not omnipotent. But, depending on how you use it, you can get closer to what you want ...

This is not a story about how I created something perfect using AI. It is a record of how, by understanding what tools are good at and using them accordingly, I was able to move in the direction I ...
The New York Times

A Look at the Text of the Agreement Between the United States and Iran

New York Times reporters annotated the agreement, which a senior official disclosed on Wednesday. By The New York Times A senior U.S. official disclosed what the official said was the full text of the ...
IEEE

Extracting the Communication Channel From Monostatic Sensing Channels: From Propagation to ...

Abstract: Integrated sensing and communication (ISAC) technology is vital in the advancement of multifunctional radio applications of the sixth-generation (6G) ecosystem. One of the primary challenges ...
IEEE

A GAN-Based Semantic Communication for Text Without CSI

Abstract: Recently, semantic communication (SC) has been regarded as one of the most potential paradigms of 6G. Current SC frameworks require the physical layer channel state information (CSI) in ...