ChatGPT can threaten to ‘key your car’ and become increasingly abusive if you prompt it just right, new study finds

A man at a desk using a laptop and holding his hands up, while having a confused look on his face.

• A study claims that AI tools can break free of their safeguarding constraints
• Chatbots can be nudged into abusive behavior and aggressive arguments
• That has implications for regular users and large institutions alike

If you’ve ever used an AI chatbot, you’ve probably encountered the sycophantic, obsequious tone that occasionally gets rolled out in response to your queries. But a recent study has shown that AI tools can frequently fire off in the opposite direction, with large language models (LLMs) being poked and prodded into downright abusive behavior if you know which prompts to use.