New Study Reveals Human Conversations Increasingly Mirror AI Responses

By Abdul Wasay ⏐ 2 weeks ago ⏐

2 min read

Study Shows That People Talk A Lot Like Ai Chatgpt Now

A new study has set off alarm bells by showing that ChatGPT and other AI are significantly altering day-to-day vernacular in a major language shift.

Researchers tracked thousands of Reddit posts, forum comments, and text messages. They found rising use of phrases, sentence structures, and filler words that match ChatGPT’s trademark style.

Study Reveals the Growing ChatGPT Language Shift

The team at Germany’s Max Planck Institute compared writing samples from 2022 and 2025. They discovered a steep jump in polite hedges such as “as an AI language model,” plus an uptick in formal connectors like “moreover” and “consequently.” Scientists attribute this to users copying and pasting AI text or unconsciously imitating it during chats.

A parallel MIT experiment offered similar proof. Volunteers who drafted emails with ChatGPT later wrote independent messages that echoed the bot’s cadence and vocabulary, even when told not to. Brain-scan data showed reduced creative activity, hinting at long-term effects on cognition.

Experts Warn of Unintended ChatGPT Language Shift Consequences

Linguists fear the trend could flatten cultural nuance. Oxford researchers note that AI favors global English with neutral slang, eroding local idioms and minority dialects. Over time, a single “AI accent” might dominate online spaces, suppressing diverse expression.

Tech ethicists add another worry. If the public adopts AI phrasing wholesale, future models will train on those AI-like texts, creating a feedback loop that amplifies blandness. One researcher likened it to “photocopying a photocopy until the picture fades.”