AIPhys.org Technologyabout 3 hours ago
Study Finds AI Chatbots Unsafe for Abuse Survivors
3 min read
A study led by NYU Tandon researchers presented in August 2026 reveals that 86% of AI chatbot responses to survivors of tech-enabled abuse lack accuracy or safety. The team tested four systems, including Aimee and Ruth, finding 60% omitted critical warnings. These findings were shared at the USENIX Security Symposium in Baltimore, highlighting severe risks in automated safety advice.
Level
Hype check
Tap to vote and see what everyone thinks.
