Researchers from Boston University, Northeastern University, KAIST, and University of Murcia, et al. have released “FHECore: Rethinking GPU Microarchitecture for Fully Homomorphic Encryption”.

AI智能体正颠覆反欺诈格局。本文剖析了攻防双方的“非对称竞争”，指出智能体网络能以群体智能、数据协同和毫秒级响应扭转劣势。同时强调，破解数据隐私与AI自身安全是落地关键。最终预言：反欺诈行业将进化为“智能体安全”，让信任成为业务的内生基因。

Enterprises seeking to make good on the promise of agentic AI will need a platform for building, wrangling, and monitoring AI agents in purposeful workflows. In this quickly evolving space, myriad ...

Octra Network deploys on-chain FHE machine learning with governance and zero-knowledge verification, letting anyone run private ML inference directly on-chain.

Andes Y. L. Kei, Sherman S. M. Chow PAPER SHAFT: Secure, Handy, Accurate and Fast Transformer Inference Adoption of transformer-based machine learning models is growing, raising concerns about ...

【新智元导读】刚刚，YC最新创业清单点名「AI欺诈猎手」。当黑灰产开始用AI作案，防守方也在组建智能体军团——反欺诈的终局，或许不是更强风控，而是一个自带安全基因的智能体世界。

With all of the talk about the security risks of AI, one issue that seems to be overlooked is this: the fact that AI systems only function by exposing their most valuable assets — models and data.

Indian inventor and system theorist Vatsal Soin has recently filed patent US 19/489,595, India 202511115781, and ...

South Korean Telecoms Compete in AI Race at MWC 2026 SK Telecom, KT, LG Uplus showcase Full-Stack AI, K-culture integration, ...

Every AI-mediated decision in commerce — a product match, a logistics authorisation, a supplier selection, a real-time delivery assignment, a food security coordination — generates a cryptographic ...