Please turn JavaScript on
Find more feeds
8kSec icon

8kSec

Follow 8kSec's news and updates in a matter of seconds! We will deliver any update via email, phone or you can read them from here on the site on your own news page.

You can even combine different feeds with the feed for 8kSec.

Subscribing and unsubscribing is fast, easy and risk free.

The whole service is free of cost.

8kSec: 8kSec | Live & On-Demand Mobile & AI Security Trainings

Is this your feed? Claim it!

Publisher:  Unclaimed!
Message frequency:  0.18 / day

Message History

Red Teaming LLM Applications at Scale with garak and PyRIT
How to move from ad-hoc prompt fiddling to a repeatable AI red teaming pipeline using garak, PyRIT, Promptfoo and DeepTeam — and how to wire continuous LLM security testing into CI/CD.

Read full story
AI-Assisted Fuzzing: Generating Harnesses with a Local LLM
A hands-on guide to using a local open-source LLM to write fuzzing harnesses, then finding a real memory-corruption bug with AddressSanitizer — with runnable code and a downloadable lab.

Read full story
Poisoning the Well: RAG Knowledge-Base Attacks and How to Defend Them
Retrieval-Augmented Generation trusts whatever the retriever returns. This post breaks down PoisonedRAG-style knowledge-base poisoning, embedding attacks, and the provenance, filtering, and verification defenses that stop them.

Read full story
Reverse Engineering with a Local LLM: Disassembly, Triage, and Frida Hooks
Use a local open-source LLM to explain ARM64 disassembly, recover a license-check algorithm, and generate a working Frida bypass — reproducibly, with a downloadable lab.

Read full story
Attacking the Model Context Protocol: Tool Poisoning, Rug-Pulls, and CVE-2025-6514
How the Model Context Protocol expands the agent attack surface — tool poisoning, rug-pull attacks, and the mcp-remote command injection (CVE-2025-6514, CVSS 9.6) — plus how to build MCP servers and clients that survive a hostile counterpart.

Read full story