groundy

security

  1. jun 04securityChatGPT's New Lockdown Mode Borrows Apple's Name for a Prompt-Injection Kill Switch
  2. jun 04securityStored Prompt Injection Now Persists Across AI Agent Sessions
  3. jun 04securityLLM Data Poisoning Survives the Data-Cleaning Defenses Built to Stop It
  4. jun 01securityWhy Attack Success Rate Misleads LLM Jailbreak Benchmarks
  5. may 27securityOpenAI's New Safety Bug Bounty Pays Researchers for Jailbreaks and Policy Bypasses
  6. may 23securityAI Jailbreaks Are Now a Reasoning Problem, Not a Prompt Problem
  7. may 18securityTrustFall: One Keypress in Claude Code, Gemini CLI, Cursor, and Copilot CLI Triggers Unsandboxed RCE
  8. may 18securityMultiBreak Benchmark: 10,389 Multi-Turn Jailbreak Prompts Raise ASR 54pp on DeepSeek-R1-7B
  9. apr 29securityInstructLab CVE-2026-6859: Hardcoded trust_remote_code=True Turns Any HuggingFace Model Into RCE
  10. apr 29securityMercor's 4TB Lapsus$ Breach Hands Voice-Clone Attackers 40,000 Pre-Verified Targets
  11. apr 24securityCitizen Lab's 'Bad Connection' Names Three Telecom Entry Points, Shows Diameter Silently Falls Back to SS7
  12. apr 23securityMarch-April MCP CVEs Expose the Local-Host Trust Model in AI Agent Frameworks
  13. feb 20securityThe Mysterious Case of Chinese Bot Traffic in 2026: How AI-Powered Bots Are Rewriting the Rules of Detection