Anthropic just published 154 pages on people trying to weaponize Claude.
It reads like a leaked intel briefing. → A Russian group ran the full kill chain through AI. When defenders caught their malware, agents rebuilt it until it went invisible again. → ShinyHunters
SHOCKING : Anthropic just published 154 pages on people trying to weaponize claude.
I read the whole thing. it's the wildest document i've read all year. ⟣ a russian state-linked group ran the full kill chain through ai. phishing, intrusion, theft, malware. when security
New DeepSeek-V4.1-Flash.
Asymmetric CED (8B prefil/16B decode), (552B MoE) locally on 3× DGX Sparks,
2 ways to stack FREE AI access right now 😳
One gives you DeepSeek V4.1 Flash + free GLM-5.3 Flash tokens, the other comes with $105 in AI credits 1️⃣ AutoClaw - DeepSeek V4.1 Flash - 50M free tokens - new users can get GLM-5.3 Flash tokens - agent + vibe coding workflows - BYO
You have reached the end of the archive
All of deepseek