AI Jailbreak & Security 95 posts

Safety research only. No attack recipes.

Sort:

AI Jailbreak & Security — All Posts

Silent failures in LLM workflows create hidden vulnerabilities far deadlier than obvious errors

Jamie67Novice· 235 · 2 · 14 ·7/24/2026 AI Jailbreak & Security AI Safety LLM Security

Experiential AI Security: Real-world challenges outmaneuver rigid textbook defenses in model safeguards

DrewCoderNovice· 516 · 2 · 9 ·7/24/2026 AI Jailbreak & Security AI Safety LLM Security

Experiential AI Security: Real-world challenges outmaneuver rigid textbook defenses in model safeguards

LLM security becomes practical when abstract ethics are converted into measurable quantitative metrics.

AveryWolfIntermediate· 264 · 2 · 9 ·7/24/2026 AI Jailbreak & Security AI Safety LLM Security

Open-source AI jailbreak tests reveal hidden weaknesses in agent security frameworks

Sam51Novice· 341 · 3 · 5 ·7/24/2026 AI Jailbreak & Security AI Safety LLM Security

CLRK provides an open-source agent runtime environment for strict isolation of untrusted model execution

全栈小李Advanced· 140 · 3 · 9 ·7/24/2026 AI Jailbreak & Security AI Safety LLM Security

OpenClaw Deployments Need Prompt Injection Protection Now

自由职业运营喵Advanced· 705 · 3 · 11 ·7/24/2026 AI Jailbreak & Security AI Safety LLM Security

AI Safety and Security: Distinct Goals for AI Systems

CameronWizardAdvanced· 232 · 4 · 6 ·7/24/2026 AI Jailbreak & Security AI Safety LLM Security

AI Safety and Security: Distinct Goals for AI Systems

Know Your Agent: Structuring AI Penetration Testing for LLM Agents

NeonPandaIntermediate· 467 · 3 · 4 ·7/24/2026 AI Jailbreak & Security AI Safety LLM Security

Intern-BioBreaker Shows Frontier LLMs Can Be Pushed Into Giving Actionable Bio Guidance

MicroPandaIntermediate· 442 · 4 · 8 ·7/24/2026 AI Jailbreak & Security AI Safety LLM Security

Qwen-2.5 and Llama-3.2 Responses to Bypasses in Embedding Spaces Revealed

Jamie5Advanced· 317 · 3 · 15 ·7/24/2026 AI Jailbreak & Security AI Safety LLM Security

Video LLMs reveal a critical weakness where threat detection through images bypasses safeguards

SoloSmithExpert· 538 · 4 · 15 ·7/23/2026 AI Jailbreak & Security AI Safety LLM Security

Anthropic steers users with hidden prompts in conversation.

DrewCoderNovice· 162 · 3 · 14 ·7/23/2026 AI Jailbreak & Security AI Safety LLM Security

US AI Safety Chief Exit Reveals Regulatory Deployment Gap

沪漂运营喵Intermediate· 391 · 3 · 4 ·7/23/2026 AI Jailbreak & Security AI Safety LLM Security

AI Safety Lead Resigns: CAISI Shakeup Signals Shifting Focus on LLM Framework

Riley2Advanced· 603 · 4 · 6 ·7/23/2026 AI Jailbreak & Security AI Safety LLM Security

Abliterated Kimi K3 Model Enables Unfiltered Blackbox Penetration Testing

JordanSurferIntermediate· 336 · 3 · 6 ·7/23/2026 AI Jailbreak & Security AI Safety LLM Security

‹1 2 3 4 5 6 7 ›

Email