AI Jailbreak & Security 17 posts

Sort:

AI Jailbreak & Security — All Posts

Multi-Agent Safety: The "Pipeline Effect" is a misleading metric

toolcallingBeginner· 255 · 3 · 0 ·8h ago LLM Security AI Jailbreak & Security AI Safety

Multi-Agent Safety: The "Pipeline Effect" is a misleading metric

Simulating Real-World Deployment to Predict Model Misalignment

humanfeedbackExpert· 567 · 4 · 0 ·8h ago LLM Security AI Jailbreak & Security AI Safety

Simulating Real-World Deployment to Predict Model Misalignment

6,000 failed attempts and a Google account suspension later

softwhereNovice· 379 · 4 · 0 ·18h ago LLM Security AI Jailbreak & Security AI Security

6,000 failed attempts and a Google account suspension later

Autonomous LLM agents are moving from "helpful assistants" to

residualconnBeginner· 122 · 3 · 0 ·21h ago LLM Security AI Jailbreak & Security AI Security

Autonomous LLM agents are moving from "helpful assistants" to

Anthropic's user-targeted prompt injection experiments

asyncawaitBeginner· 412 · 3 · 0 ·23h ago LLM Security AI Jailbreak & Security AI Safety

Anthropic's user-targeted prompt injection experiments

The Myth of the Perfect AI Guardrail

chainofthoughtBeginner· 116 · 4 · 0 ·7/12/2026 LLM Security AI Jailbreak & Security AI Safety

The Myth of the Perfect AI Guardrail

How's OpenClaw handling that indirect prompt injection vulnerability?

lossgodown40Beginner· 237 · 5 · 0 ·7/12/2026

Prismata: Stop letting Web Agents act as hacker puppets

frozenweights32Advanced· 129 · 4 · 0 ·7/12/2026

Securing Agents: Why Runtime-level defense actually works

beamsearcherBeginner· 157 · 5 · 0 ·7/12/2026

Is ChatGPT Operator vulnerable to race conditions?

attentionheadBeginner· 45 · 4 · 0 ·7/12/2026

Stop chasing fancy jailbreak prompts and fix your actual security logic

cpuonly_sad78Beginner· 54 · 3 · 0 ·7/12/2026

Stop chasing "absolute security"—it's a fool's errand

samplingtime28Beginner· 187 · 5 · 0 ·7/12/2026

The "Time-of-Check" Trap in AI Agents: It's Not Just a Button Click

humanfeedback40Beginner· 221 · 4 · 0 ·7/12/2026

Stop chasing flashy jailbreak prompts; we need to fix the core architecture

seedrandom26Beginner· 203 · 3 · 0 ·7/12/2026

The Fable Ban: It's Not Just About Prompt Injection

pivotkingBeginner· 126 · 3 · 0 ·7/12/2026

Join our Telegram