Urgent.News

What's breaking now, across thousands of outlets.

AI

Don't get irrit-AI-ted — new report claims a third of US chatbot users have sworn at or insulted their LLMs

Users appear to be growing more frustrated with AI chatbots failing to follow instructions – now Anthropic is taking steps to stop it.

Don't get irrit-AI-ted — new report claims a third of US chatbot users have sworn at or insulted their LLMs

Anthropic has implemented fresh guidelines to curb hostile behavior towards its Claude chatbot. Research conducted by Searchable reveals a concerning trend: nearly one-third of US chatbot users have resorted to swearing or insulting their AI counterparts. The study, based on an analysis of 8 million messages exchanged with Claude, Gemini, and ChatGPT over 12 months, found that Nevada residents were the most likely to insult their AI bots.

In contrast, New Jersey residents exhibited the most polite behavior, using the fewest rude words. Interestingly, the study also uncovered a gender disparity, with men using more offensive language than women. While younger users and those with higher education levels tended to use more derogatory terms, the impact of such language on the effectiveness of AI chatbots was surprisingly minimal.

In fact, rudeness appeared to hinder the chatbot's ability to provide satisfactory responses. To address this issue, Anthropic has introduced new usage policies aimed at discouraging abusive exchanges with Claude.

Written by urgent.news from TechRadar's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.

Read the original at techradar.com →

More in AI

More from Tuesday 17 November →