SecurityDigital Trendsabout 1 hour ago
NYU Study Exposes AI Chatbot Failures
3 min read
A new NYU Tandon study tested four AI chatbots on questions about stalking and found experts rated 86% of replies inaccurate, incomplete, or unsafe. Unsafe answers ranged from 55% for GPT-4o to 74% for Aimee, with Claude at 65%. Researchers put 193 real-world questions to the bots, including scenarios about finding spyware and adversarial prompts for surveillance access.
Level
Hype check
Tap to vote and see what everyone thinks.
