AI chatbots leave subtle marks on everyday language, linguists find
Northeastern University researchers say AI chatbots are leaving a linguistic fingerprint on everyday writing, with words like 'delve' and 'insight' becoming more common. They note that language change is natural and that AI both influences and assists human expression. The experts caution that distinguishing AI's influence from natural evolution is difficult.
The research draws on observations from three Northeastern University specialists who note that language evolution is a constant process, citing the rapid emergence of pandemic-era terms like "WFH" and "lockdown" as recent examples. They distinguish between vocabulary shifts, which tend to be fleeting, and grammatical changes, which persist across generations. The experts point to AI's tendency to default to English-style subject-verb-object sentence structures even when generating text in languages like Russian that permit flexible word order, potentially eroding subtle meaning conveyed through arrangement.
The researchers identify specific markers of AI-influenced writing, including elevated use of words such as "delve" and "essential," along with recurring metaphors involving tapestries and symphonies. They emphasize that chatbots both shape human expression and serve as tools for it, making it difficult to separate machine influence from organic linguistic drift.
This research could affect how people assess authenticity in writing, from academic work to casual online posts. Readers may become more attuned to AI-influenced phrasing, potentially shifting expectations for what constitutes original expression. The flattening of grammatical flexibility across languages could gradually reduce cultural and linguistic diversity, though such changes typically unfold over generations. Writers and educators may need to consider how AI-assisted tools shape communication habits without necessarily diminishing human creativity.