A new study has revealed that despite widespread public discussion, artificial intelligence tools are used in less than one percent of daily online activity for most individuals. The research, published in *Cyberpsychology, Behavior, and Social Networking*, analyzed over 14 million website visits to provide objective data on AI adoption rates. It also identified a correlation between more frequent AI use and certain personality traits.
Key Takeaways
- AI-related browsing accounts for less than 1% of total online activity for most people.
- ChatGPT is the dominant AI platform, representing over 85% of all AI-related visits.
- Frequent AI users, especially students, show higher scores in traits like Machiavellianism, narcissism, and psychopathy.
- AI is often used as a productivity tool for academic or work-related tasks.
- Self-reported AI use often does not match actual browsing data.
Measuring Actual AI Engagement
Public and academic interest in artificial intelligence continues to grow. Discussions often focus on AI's expanding role in various sectors, including education, work, and entertainment. However, much of the existing data on AI use comes from self-reported surveys. These surveys can be unreliable, as individuals may misjudge or misremember their actual technology usage, particularly with newer tools.
To overcome the limitations of self-reported data, researchers aimed to directly measure AI engagement. Their goal was to quantify how often AI tools are used in everyday life. They also sought to identify the typical user profiles and understand how AI integrates into broader online activities. This objective approach provides a clearer picture of real-world AI adoption.
"There’s been enormous public discussion about AI and its societal impact, but surprisingly little objective data on how people actually use it in their everyday browsing," stated Emily McKinley, a PhD candidate at the University of California, Davis, and lead author of the study. "Despite widespread concerns and excitement about tools like ChatGPT, we had almost no baseline understanding of actual usage patterns."
Study Design and Data Collection
The research involved two distinct studies. The first included 499 university students. The second focused on 455 members of the general public. Participants in both groups provided their web browsing history for up to 90 days. Data collection was limited to Google Chrome users due to specific data export requirements.
In addition to browsing data, participants completed surveys. These surveys measured personality traits, attitudes toward AI, and demographic information. Researchers identified AI-related visits by cross-referencing browsing history with a list of popular AI websites, such as ChatGPT and Microsoft Copilot. A large language model then categorized other website content. This allowed for analysis of AI visit frequency, adjacent browsing activities, and correlations with psychological traits.
Key Statistic: AI Browsing
- In the student sample, AI use constituted approximately 1% of all website visits.
- For the general public, AI use was even lower, averaging just 0.44% of total website visits.
- ChatGPT accounted for over 85% of all identified AI-related browsing.
AI Usage Patterns and User Profiles
The findings showed consistently low rates of AI usage. Among students, AI use made up an average of only 1 percent of all website visits. Most students rarely used AI tools. A small group of users generated the majority of AI-related traffic. ChatGPT was by far the most visited AI site, accounting for over 85 percent of all AI interactions.
While AI use among students surpassed visits to web versions of some social media platforms like Instagram, it remained significantly lower than search engine usage. "We were genuinely surprised by how infrequent AI use was, even among students who typically serve as early adopters of emerging technologies," McKinley noted.
Personality Traits Linked to AI Use
The study also explored the psychological profiles of AI users. Students who used AI more often showed slightly higher scores on personality traits associated with narcissism and psychopathy. These individuals also generally held more positive views about AI. Weak links were found between AI use and demographics like income and gender. However, age and ethnicity did not show significant relationships with AI use in this student group.
For the general public sample, AI usage rates were even lower, at an average of just 0.44 percent of website visits. ChatGPT again dominated as the most visited AI platform. In this group, fewer strong correlations between personality and AI use were observed. However, a modest relationship existed between Machiavellianism and AI browsing. Individuals with more favorable views of AI were also somewhat more likely to use it.
Understanding Aversive Personality Traits
The study refers to "aversive personality traits" such as Machiavellianism, narcissism, and psychopathy. These traits are part of the 'Dark Triad' in psychology. Machiavellianism involves manipulation and a cynical disregard for morality. Narcissism is characterized by grandiosity, pride, egotism, and a lack of empathy. Psychopathy is marked by antisocial behavior, impulsivity, selfishness, and a lack of remorse.
Researchers specifically identified "prolific users" as those whose AI browsing exceeded 4 percent of their total website visits. Among students, these prolific users scored significantly higher on measures of Machiavellianism, narcissism, and psychopathy compared to their peers. These patterns were less clear in the general public sample, possibly due to the lower overall AI usage in that group.
"Interestingly, people who use AI more tend to score higher on aversive personality traits, particularly Machiavellianism, narcissism, and psychopathy, though these patterns were stronger among students," McKinley explained.
AI as a Productivity Tool
The study also examined online activities immediately before and after AI use. Many participants visited internet and telecom sites, such as search engines or login pages, before using an AI website. After AI use, participants were more likely to visit websites related to education, computers, or professional tasks. This pattern suggests that AI tools are frequently integrated into workflows, particularly in academic or job-related contexts. The researchers propose this indicates AI is perceived more as a productivity tool than a source of entertainment.
Another important finding concerned the discrepancy between self-reported and actual AI use. In the general public sample, participants were asked to estimate their AI usage frequency. The correlation between these estimates and their actual browsing data was only moderate. This indicates that self-reports alone are not a reliable method for assessing AI use, aligning with prior research on other forms of media consumption.
- Before AI use: Common activities included search engines and login pages.
- After AI use: Common activities included education, computer, or professional task-related websites.
Limitations and Future Research Directions
While this study offers valuable objective data on AI use, it has some limitations. The analysis focused solely on web-based interactions, excluding mobile app usage, which may be significant for some users. Additionally, only Google Chrome users participated, potentially influencing the sample's representativeness.
The research primarily examined a narrow set of individual traits, such as personality and basic demographics. Other factors like emotional well-being, specific motivations, or social environments could also impact AI adoption but were not included in this study. The research team suggests that future investigations should explore how these additional psychological and social variables relate to AI use.
Understanding the *content* of interactions on AI platforms is also crucial. This study captured how often users visited AI websites, but it could not determine whether they were writing essays, solving problems, or exploring out of curiosity. Tracking the specific content of these interactions would provide deeper insights into user goals and intentions.
As AI becomes more integrated into daily life, usage patterns are likely to evolve. The researchers plan to continue this line of inquiry, examining whether AI use is linked to specific outcomes. "We want to understand not just how often people use AI, but what they’re using it for and how that content relates to their individual characteristics," McKinley elaborated. "We’re also interested in examining the downstream consequences of AI use: for example, does actual usage predict outcomes like academic integrity, information-seeking behaviors, or work performance?"
This study represents a significant step in objectively quantifying AI use in naturalistic settings. It utilized months of actual browsing data, moving beyond the limitations of self-report. The methodological approach, combining passive digital trace data with psychological measures, offers a template for future technology research. This can help improve understanding of how individual differences shape technology adoption in the real world.
The study, titled "Evaluating Artificial Intelligence Use and Its Psychological Correlates via Months of Web-Browsing Data," was authored by Emily McKinley, David M. Markowitz, Rui Zhu, and Brandon Van Der Heide.