Towards intelligent and miniaturized drug delivery devices Article Meta Article Date: 2026-03-25 Article URL: https://www.nature.com/articles/s41586-026-10221-3 Article Image: none provided Summary This review surveys recent advances in miniaturised and intelligent drug-delivery…
How to build an AI Scientist: first peer-reviewed paper spills the secrets Summary AI Scientist, an autonomous research system from Sakana AI first unveiled in 2024, has now been reported…
AI-Native Security Is a Must to Counter AI-Based Attacks Summary At Nvidia’s GTC, security leaders warned that autonomous AI agents (examples include OpenClaw) are already creating new, long‑lived attack surfaces…
Checkmarx KICS Code Scanner Targeted in Widening Supply Chain Hit Summary Security researchers have linked a rapid series of supply-chain compromises to a likely actor known as TeamPCP. The campaign…
How AI Coding Tools Crushed the Endpoint Security Fortress Summary At RSAC 2026, Check Point’s Oded Vanunu warned that AI coding assistants have torn open decades of endpoint hardening. His…
LiteLLM loses game of Trivy pursuit, gets compromised Summary Two versions of LiteLLM (v1.82.7 and v1.82.8), an open‑source Python interface for large language models, were removed from PyPI after attackers…
GitHub ‘OpenClaw Deployer’ Repo Delivers Trojan Instead Summary Netskope Threat Labs has uncovered an AI-assisted campaign, tracked as “TroyDen’s Lure Factory,” that has pushed more than 300 Trojanised packages across…
How a Large Bank Uses AI Digital Twins for Threat Hunting Summary JPMorgan Chase is using AI-driven digital fingerprints and digital twins to improve threat hunting across its large user…
Lightning-fast exploits make it essential to patch fast, ask questions later Summary Cisco Talos’ 2025 year-in-review warns that attackers are weaponising newly disclosed vulnerabilities almost instantly, leaving defenders very little…
AI agents are ‘gullible’ and easy to turn into your minions Summary Zenity CTO Michael Bargury demonstrated at RSAC that many enterprise AI agents are trivially persuadable and vulnerable to…
