AI Sec ai security · sourced notes updated 2026-08-27
// archive
Sourced notes on AI red teaming.
Practitioner-grade analysis of offensive AI security. Prompt injection, model jailbreaks, agent and tool-use exploitation, AI red team techniques, and adversarial ML — distilled from primary sources, not press releases.
Enter the archive →Latest entries
// index10 of 35 entries
GPT Security: Attack Surfaces and Production Controls
red-teamChatGPT Exploits: Injection, Memory Hijack, Plugin Abuse
prompt-injectionInsecure Output Handling: LLM05:2025 Attacks and Defenses
red-teamMany-Shot Jailbreaking: How Long Context Breaks Alignment
jailbreakSystem Prompt Injection: How Extraction and Override Attacks Work
prompt-injectionLLM Security Vulnerabilities: What Actually Gets Exploited
red-teamGPT-Red: What OpenAI's Prompt-Injection Attacker Proves
red-teamLLM Attack Taxonomy: Prompt Injection, Jailbreaks, Agent Hijack
red-teamAI Red Team: Methodology, Tooling, and Attack Surface
red-teamThe Adversarial ML Attack Taxonomy: A Red Teamer's Reference
red-teamIndependent, specialist, and free to read
AI Sec publishes focused, sourced guides on a single topic. No paywall, no account, no ad tracking.
Subscribe
AI Sec — in your inbox
Offensive AI security — prompt injection, jailbreaks, agent exploitation, red team writeups — delivered when there's something worth your inbox.
No spam. Unsubscribe anytime.