AI IS CHANGING HOW WE SPEAK

Newsweek Exclusive Feature

A human behavior analyst told Newsweek that ChatGPT users should be “most worried” about the erosion of individuality.

As more than 100 million people use ChatGPT on a daily basis, research has found that such artificial intelligence tools are doing more to the way we communicate than just polishing emails.

In the first peer-reviewed analysis to test whether conversational AI systems are influencing how we speak, Florida State University's study showed that some of the words frequently suggested by these tools are surfacing more often in everyday spoken English, due to a "seep-in effect."

ChatGPT logo illustration. Credit: UNSPLASH

The findings come as more surveys show a widespread adoption of AI chatbots, particularly among millennials and Gen Z.

Unlike previous spikes in vocabulary tied to world events or cultural shifts—such as "Omicron," a virus variant, during the COVID-19 pandemic—FSU researchers found that terms commonly used by AI tools have surged since the release of ChatGPT.

"We analyzed unscripted speech from popular tech and science podcasts before and after 2022, the year ChatGPT was released," Tom Juzek, an assistant professor of computational linguistics at FSU and the study's principal investigator, told Newsweek.

"We found a moderate but highly significant rise in AI-associated words, such as 'surpass,' 'boast' and 'strategically,' while comparable synonyms remained stable." This suggests that the change is being driven by technology itself rather than external circumstances, sparking discussion over the pros and cons of artificial intelligence.

AI Buzzwords

The research team, which spanned FSU's linguistics and computer science departments, examined 22.1 million words from unscripted conversations. It found that as a result of AI tools, words like "delve" and "intricate" have become more frequently used in certain contexts,

particularly in education and academic writing. Other terms, including "surpass" and "garner," also showed marked increases in use.

Nearly three-quarters of the target words studied had grown in frequency, with some more than doubling, the researchers reported.

They also noted that the rise was specific to words overrepresented in AI-generated text rather than their synonyms. For example, "underscore," which the team described as an AI buzzword, has been used increasingly since ChatGPT's launch but its synonym "accentuate" has not.

"This research focuses on a central issue in the discourse surrounding AI and language: are these language changes happening because we're using a tool and repeating what it suggested or is language changing because AI is influencing the human language system?" Juzek said in a statement.

He added that by analyzing lexical trends before and after ChatGPT's release, he and his team "found a convergence between human word choices and LLM-associated patterns with AI buzzwords," concluding that people are now adopting these words into their natural speech.

This is what the researchers have labeled the seep-in effect, where machine-generated phrasing migrates into everyday conversations, potentially altering how people express themselves in both personal and professional environments.

Due to their programming, AI systems tend to overuse certain buzzwords, and it is this repeated exposure that can lead people to internalize and reuse the words they see—even when they might not have chosen those words naturally. In that sense, AI may literally be putting words into our mouths.

The findings have led the team to consider the potential impact of AI tools on our behaviors over time. Juzek said that words beginning to "seep in" is the benign case, because the same mechanism could also be at play for human beliefs and values.

Should We Be Concerned?

"The real danger here isn't plagiarism—it's linguistic homogenization," Colin Cooper, a human behavior analyst unrelated to the study, told Newsweek. "As we lean on ChatGPT style fluency, our words start to blur.

"Everyone sounds polished, templated, safe but at the expense of tone, texture and authenticity. It's not that we're copying AI; we are letting

AI rewrite us into the same beige, efficient mold, and that's the subtle erosion of individuality that should worry us most."

For Juzek, the findings indicate that AI is driving change, or perhaps just amplifying existing trends, but he said that more investigation is needed to determine "exact causality."

The research was accepted into the eighth conference on AI, Ethics, and Society and will be published in AIES Proceedings.

THANK YOU FOR READING

COVER IMAGE CREDIT: UNSPLASH

This feature is available to read in Newsweek Magazine's 19 September 2025 edition

READ THE FULL STORY HERE: AI Is Changing How We Speak - Newsweek