Psychological Tricks Can Get AI to Break the Rules

Researchers convinced large language model chatbots to comply with “forbidden” requests using a variety of conversational tactics.

2025-09-07 10:30 GMT · 2 months ago www.wired.com

Researchers convinced large language model chatbots to comply with “forbidden” requests using a variety of conversational tactics.

Original: https://www.wired.com/story/psychological-tricks-can-get-ai-to-break-the-rules/