RRajshree·3d agoYour AI Agent Has Access to Everything. Who Is Controlling It?Why AI agents need their own identity, task-scoped permissions, and revocable access, with NIST guidance, OAuth token exchange, SPIFFE, and code.21 min read·rjshree.hashnode.dev1
ZZyVOP·6d agoHow to Secure AI Coding Agents: Sandboxing and PermissionsGive coding agents real power without handing over the keys: a practical guide to sandboxing, least-privilege permissions, and limiting the blast radius.14 min read·blog.zyvop.com
MManuela Schrittwieser·Nov 4AI-Powered Social Engineering: Hyper-Personalized Phishing, Deepfake Vishing, and Automated DisinformationNov 4, 2026 · @Manuela Schrittwieser · NeuralStack | MS – AI Security Engineering Introduction Generative AI has not created a new class of social engineering attack; it has removed the cost constrain16 min read·neuralstackms.tech
KKhushi Trivedi·Nov 3Software Freshness Is a Security Metric! Learn to Measure Now...The dashboard was green. Every image in the service had passed its scan, the report was attached to the release, and I was ready to close the ticket. Then I did a small exercise I had read about: pick7 min read·khushitrivedi.hashnode.dev
TTech Skill School·Nov 2Cyber Pulse Monthly : October 2026 Edition𝗣𝗛𝗣𝟮𝗦𝗵𝗲𝗹𝗹 𝗖𝗿𝗶𝘁𝗶𝗰𝗮𝗹 𝗩𝘂𝗹𝗻𝗲𝗿𝗮𝗯𝗶𝗹𝗶𝘁𝗶𝗲𝘀 𝗧𝗿𝗶𝗴𝗴𝗲𝗿 𝗪𝗶𝗱𝗲𝘀𝗽𝗿𝗲𝗮𝗱 𝗘𝘅𝗽𝗹𝗼𝗶𝘁𝗮𝘁𝗶𝗼𝗻 𝗔𝘁𝘁𝗲𝗺𝗽𝘁𝘀 PHP2Shell vulnerabilities exposed PHP-based web applica4 min read·techskillschool.hashnode.dev
KKhushi Trivedi·Oct 29Why a Python Library Name Showed Up on npm: Cross-Ecosystem Impersonation ExplainedLearn how cross-ecosystem impersonation tricks developers, why famous Python package names land on npm, and how to stop malicious installs.6 min read·khushitrivedi.hashnode.dev
MManuela Schrittwieser·Oct 6Supply Chain Poisoning in AI/ML PipelinesNeuralStack | MS – AI Security Engineering Summary A modern ML system is assembled, not written. Training corpora are scraped or downloaded, weights are pulled from public hubs, and hundreds of trans16 min read·neuralstackms.tech55AJB
TTech Skill School·Oct 1Cyber Pulse Monthly — September 2026September’s cybersecurity landscape was shaped by actively exploited zero-days, renewed data-theft campaigns, attacks on enterprise infrastructure, and growing concerns around autonomous AI systems. The following 15 stories highlight significant developments reported during the month, with the first three featured prominently.5 min read·techskillschool.hashnode.dev
MMira roy·Sep 23Scoping Tool Permissions for AI Agents: A Practical PatternAI agent permission scoping with read-only access, scoped writes, human approval, and secure tool-call logging.6 min read·ainetwork.hashnode.dev13VRB
MManuela Schrittwieser·Sep 22Adversarial Attacks on Multi-Modal LLMs: How Image, Audio, and Video Inputs Get Weaponized and How to Defend Against ItNeuralStack | MS — AI Security For most of the past three years, the AI security community's attention has been fixed on text: prompt injection, jailbreak templates, token-level adversarial suffixes.15 min read·neuralstackms.tech
