​A study by Stony Brook University and the University of Pennsylvania, published in the Findings of the Association for Computational Linguistics on November 29, 2021, analyzed over three million tweets from 3,000 social bot accounts and an equal number of genuine accounts. The research estimated 17 human attributes—such as age, gender, five personality traits, eight emotions, and sentiment—for each account. While individual bots appeared human-like, their uniformity across these attributes revealed their artificial nature, suggesting a new method for detecting bots based solely on linguistic patterns.