If you say a few words, generative AI may understand who you are—perhaps even more accurately than your closest family and friends. A new study conducted by the University of Michigan has found that widely available generative AI models, such as ChatGPT, Claude, and LLaMa, can predict personality traits, key behaviors, and daily emotions as well as, or even better than, those closest to you.
“What this study shows is AI can also help us understand ourselves better, providing insights into what makes us most human, our personalities,” said Aidan Wright, the study’s first author and a U-M professor of psychology and psychiatry. “Lots of people may find this of interest and useful. People have long been interested in understanding themselves better. Online personality questionnaires, some valid and many of dubious quality, are enormously popular.”
AI as a Judge of Personality
The research explored whether AI programs like ChatGPT and Claude can act as general “judges” of personality. To test this hypothesis, researchers had the AI analyze people’s words—either from short daily video diaries or longer recordings of their thoughts—and asked it to answer personality questions as each person would. The study included narratives from over 160 individuals collected in both real-life and lab settings.
The results were striking. The AI’s personality scores closely mirrored how individuals rated themselves, often surpassing the accuracy of ratings from friends or family. Older text-analysis methods did not perform nearly as well as these newer AI systems. “We were taken aback by just how strong these associations were, given how different these two data sources are,” Wright remarked.
Implications for Understanding Human Behavior
AI’s personality ratings could also predict tangible aspects of people’s lives, such as their emotions, stress levels, social behavior, and even whether they had been diagnosed with mental health conditions or sought treatment. This suggests that personality naturally manifests in our everyday thoughts, words, and stories, even when we are not consciously describing ourselves.
Chandra Sripada, a U-M professor of philosophy and psychiatry, noted that the findings bolster the long-held belief that language carries profound clues about psychological traits, such as personality and mood. He added that open-ended writing and speech could be a potent tool for understanding personality. Thanks to generative AI, researchers can now analyze this type of data quickly and accurately in ways that were previously impossible.
Future Directions and Questions
Despite these promising findings, important questions remain. The study relied on individuals rating their own personalities and did not compare AI assessments with judgments from friends or family, nor did it examine how results might vary across different demographics such as age, gender, or race.
Researchers also do not yet know whether AI and humans rely on the same signals or whether AI could eventually outperform self-reports in predicting major life outcomes like relationships, education, health, or career success. “The study shows that AI can reliably uncover personality traits from everyday language, pointing to a new frontier in understanding human psychology,” said Colin Vize, assistant professor of psychology at the University of Pittsburgh.
Whitney Ringwald, assistant professor of psychology at the University of Minnesota, emphasized that the results “really highlight how our personality is infused in everything we do, even down to our mundane, everyday experiences and passing thoughts.”
The study’s findings, co-authored by Johannes Eichstaedt of Stanford University and Mike Angstadt and Aman Taxali from U-M, are published in the journal Nature Human Behavior.
Study: Generative AI predicts personality traits based on open-ended narratives (DOI: 10.1038/s41562-025-02389-x)
Contact: Jared Wadley
