AITechCrunchabout 2 hours ago
AI guardrails hinder offensive cybersecurity researchers
1 min read
AI guardrails designed to block malicious hackers are also impeding legitimate offensive cybersecurity researchers. Researchers like NCC Group's Chris Anley say guardrails block key steps like asking a model to exploit a bug. Some researchers turn to open-source or Chinese models like GLM to avoid restrictions.
Level
Hype check
Tap to vote and see what everyone thinks.
