Artificial Intelligence

New Report Claims One-Third of US Chatbot Users Swear at LLMs, Prompting Policy Updates

· GlobeWire Daily Editorial Desk · Source: www.techradar.com

2 min read

Key takeawaysA recent study indicates that approximately one-third of US-based chatbot users have directed insults or profanity at large language models, prompting platform developers like Anthropic to introduce new usage policies aimed at preventing abusive interactions.
Angry man shouting at mobile phone while sitting at a desk

Image: TechRadar · Photo source · Resized and letterboxed to 16:9; no AI generation

A recent study indicates that approximately one-third of US-based chatbot users have directed insults or profanity at large language models, prompting platform developers like Anthropic to introduce new usage policies aimed at preventing abusive interactions.

According to research compiled by Searchable, which analyzed an opt-in dataset of 8 million messages exchanged with popular AI tools including Claude, Gemini, and ChatGPT, roughly 33% of US users resort to swearing or insults when communicating with chatbots. The monitoring period ran between September 2025 and September 2026, utilizing anonymized data categorized by age, education level, and geographic location.

Key Findings from the Chatbot Behavior Study

  • Regional Variations: Researchers found that Nevada recorded the highest rate of derogatory language, averaging 22.7 rude or offensive words per 10,000 words. Conversely, New Jersey emerged as the most polite state with an average of 11.4 derogatory words.
  • City Comparisons: Las Vegas recorded 22 swear words or slurs per 10,000 words, whereas New York registered 12.3 words per 10,000, making it the politest surveyed city in the US.
  • Resolution Rates: The study tracked conversational outcomes and noted that rudeness did not improve the likelihood of resolving errors. Plain corrections resulted in a 41% fix rate by the end of the conversation, compared to only 31% for sweary complaints.

Practical Implications and Policy Responses

Reports indicate that the primary driver behind uncivil language is user frustration stemming from unsatisfactory or incorrect results provided by the LLMs. However, the data suggests that maintaining composure yields better troubleshooting outcomes.

Following the release of these findings, Anthropic introduced updated usage policies specifically designed to block cruelty and abusive behavior in interactions with its Claude chatbot. Users seeking to verify original survey statistics or platform policy changes can consult official announcements directly through the respective AI providers' web portals.

Share this article
Read the original source: Don't get irrit-AI-ted — new report claims a third of US chatbot users have sworn at or insulted their LLMs

AI-assisted reporting; source claims are attributed and have not been independently verified.

About the author

GlobeWire Daily Editorial Desk — AI-assisted news based on attributed published sources.

More to read