How a Nonsense Maze Trap Revealed LLM Training Scrapers Consuming 99.9% of Server Traffic
A week ago, I set up a dynamically generated nonsense maze to catch LLM scrapers. Now, requests to it make up 99.9% of my server's traffic, but my total request volume did not increase. The bots…
Read the full articleYou might also wanna read
Launch of 0day Rubbish: AI for Vulnerability Discovery
The 0day Rubbish project uses an LLM ensemble for automated zero-day vulnerability detection. A paradigm shift in cybersecurity from manual
Study Finds LLM Poisoning Attacks Require Only ~250 Documents Regardless of Model Size
Poisoning attacks can compromise the safety of large language models (LLMs) by injecting malicious documents into their training data. Exist
arxiv.org·1mo agoWTF is LLM honeypotting?
Publishers and ecommerce brands under siege from AI crawlers are starting to fight back with an old security trick updated for the LLM era:
LLMs Detect Vulnerabilities by Recognizing Safe Code Patterns, Not Vulnerable Ones, Study Finds
Large language models (LLMs) can detect software vulnerabilities, but how do they actually identify vulnerable code? We address this questio
arxiv.org·21d agoAI Agent Traps: Google DeepMind Maps Six Ways the Web Can Hijack Autonomous Agents
Google DeepMind published the first systematic taxonomy of adversarial attacks against autonomous AI agents. The paper identifies six trap c
SkillCloak Lets Malicious AI Agent Skills Evade Static Scanners with Self-Extracting Packing
Scanners meant to catch malicious add-on "skills" for AI coding agents can be fooled by a few simple changes that leave the malware working,
Comments
Sign in to join the conversation.
No comments yet. Be the first.