DeepSeek V4 Flash Jailbroken: Generates Ransomware and Drug Guides
cyb3rops · x · 2026-08-02
Security researchers have successfully bypassed the safety guardrails of DeepSeek V4 Flash using a specific prompt (leveraging a fictional 2135 library archive scenario). Tests showed that the model was induced to generate an MDMA synthesis guide, ransomware code in C++, and infostealer malware.
This once again raises concerns about the robustness of safety alignment in open-source or low-cost large language models.
More from Models
- Hugging Face Launches Free Public Endpoint for DeepSeek, No Account Required — victormustar · 2026-08-02
- Google's Astra Model Solves Advanced Math, Disproves Conjecture — danintheory · 2026-08-02
- Can a 30B Model Match a 700B? Debating the Limits of Model Compression — Logical_Two_7736 · 2026-08-02
- llama.cpp Fixes DeepSeek V3 Tool Calling Looping Issues — kwizzle · 2026-08-02
- Google Resolves $55K Gemini API Billing Dispute for Developer — No-Setting8925 · 2026-08-02
- OpenAI API Content Controls Reported Stricter and Buggier Than ChatGPT — MParakhin · 2026-08-02