Stealing AI Reasoning: How LLM APIs Leak Secret Thinking
Zara Chen2 hours ago
LLM security
5 stories tagged LLM security.
AI. Zara Chen2 hours ago
Stealing AI Reasoning: How LLM APIs Leak Secret Thinking
Researchers found that encrypted reasoning traces in major LLM APIs can be extracted in plaintext using weaker models—exposing IP, private data, and hidden prompts.
Generative UI Looks Exciting—Until You Ask Who Controls It
Rachel "Rach" Kovacs3 months ago
AI. Rachel "Rach" Kovacs3 months ago
Generative UI Looks Exciting—Until You Ask Who Controls It
AI agents that write their own UI code are impressive. But LLM-generated code running in your browser has a trust problem most demos skip past.
OWASP's Top 10 LLM Vulnerabilities: What Can Go Wrong
Marcus Chen-Ramirez6 months ago
AI. Marcus Chen-Ramirez6 months ago
OWASP's Top 10 LLM Vulnerabilities: What Can Go Wrong
OWASP's updated Top 10 for large language models reveals how easily AI systems can be manipulated, poisoned, or tricked into leaking sensitive data.
AI Agents Need DMVs: A Reality Check on Autonomous Systems
Marcus Chen-Ramirez6 months ago
AI. Marcus Chen-Ramirez6 months ago
AI Agents Need DMVs: A Reality Check on Autonomous Systems
IBM's Jeff Crume argues AI agents need governance infrastructure like cars. But the analogy reveals more about the problem than the solution.
AI 'Skills' Are Creating a Security Nightmare
Yuki Okonkwo6 months ago
AI. Yuki Okonkwo6 months ago
AI 'Skills' Are Creating a Security Nightmare
LLM 'skills'—markdown files that enhance AI capabilities—are spreading malware, hallucinated commands, and supply chain attacks. Here's what's going wrong.