Highly conscientious people might hesitate to use generative AI models

A person’s underlying personality traits influence how often they use conversational artificial intelligence, largely depending on whether they feel the tool elevates their social status and how confident they are at operating it. Researchers recently mapped how individual psychological dispositions associate with the early adoption of text-based machine learning tools. The study was published in The Journal of Psychology.

Psychological frameworks often sort human behavior into five broad categories, known collectively as the Five Factor Model. This system breaks personality down into extraversion, openness, conscientiousness, agreeableness, and neuroticism. These traits act as a psychological baseline for human interaction. They shape how a person engages with their environment, navigates unexpected challenges, and adopts new habits throughout their lifetime.

When new digital tools emerge, behavioral researchers often look to these personality profiles to understand who will adopt the technology first. Extraversion reflects sociability and a desire to engage with the outside world. Openness involves a high capacity for curiosity, creativity, and a willingness to try novel experiences. Conscientiousness describes people who are highly organized, disciplined, and focused on practical achievements.

The Five Factor Model also includes agreeableness and neuroticism, though researchers often study these two traits separately in the context of technology. Highly agreeable individuals prioritize group harmony and typically wait for a program to become mainstream before trying it. People with high levels of neuroticism tend to overestimate the risks associated with new systems, making them hesitant to adopt unfamiliar software.

Because artificial intelligence chatbots are still relatively unpredictable, researchers focused strictly on the traits most associated with early technological adoption. Extraversion, openness, and conscientiousness generally reflect novelty-seeking behaviors, forward-thinking mentalities, and goal-directed actions. These active characteristics typically align with the profile of an early technology user attempting to gain an advantage in their daily work.

To analyze technology adoption accurately, researchers pair these internal personality traits with external behavioral frameworks. One common framework is the Technology Acceptance Model. This model evaluates how outside environmental pressures and internal technological beliefs convince a person to integrate a specific software into their routine.

Two important factors in this acceptance model are social image and computer self-efficacy. Social image refers to an individual’s personal belief that using a new technology will make them look better or more distinguished to their peers. Computer self-efficacy describes the level of confidence an individual possesses regarding their own capacity to operate the program effectively without needing outside help.

Artificial intelligence chat programs represent a newly popular category of digital utility. Because programs like ChatGPT only became publicly available in late 2022, empirical data about the people who actively use them remains scarce. Tingjun Deng and Dake Wang, researchers at Shanghai Jiao Tong University in China, designed a project to discover how intrinsic personality differences and extrinsic social perceptions associate with user habits.

In March 2023, Deng, Wang, and their colleagues surveyed 784 participants across China. The researchers limited the pool to individuals who had already tried ChatGPT at least once in their lives. The participant group consisted primarily of university students and young working professionals from eastern coastal cities.

Participants answered a series of psychological and behavioral questions using a standardized five-point scale. The survey measured the respondents’ innate levels of extraversion, openness, and conscientiousness. It also asked them to rate their personal confidence in using the chatbot, their perception of how using the software affected their social standing, and their actual frequency of use.

Following data collection, the researchers used structural equation modeling to analyze the responses. This statistical technique allows researchers to test multiple relationships between different variables simultaneously. The goal was to see if personality traits directly influenced software usage, or if social prestige and technical self-reliance acted as stepping stones between personality and behavior.

The analysis revealed that extraversion was the only measured trait with a direct, positive association with general chatbot usage. Highly extraverted people reported logging into the application more often. Extraverted individuals tend to seek out social interactions, and they might treat interactions with a responsive language model as an extension of their natural conversational habits.

The other two traits, however, influenced usage primarily through indirect pathways. Direct associations between openness and usage frequency were not statistically significant. The same lack of a direct connection was true for the relationship between conscientiousness and general usage frequency.

The researchers suggested a few potential mechanisms for these missing direct links. Highly open individuals often value imaginative, highly original ideas in their daily pursuits. Because text-generation programs output aggregated responses based on preexisting data patterns, the generated text might sometimes feel too conventional to satisfy highly creative users.

A different friction point might exist for highly conscientious individuals. Conscientious people are cautious and methodical, often preferring tools that are entirely reliable and predictable. Early versions of text-generating programs occasionally produced factual errors, which might have prompted disciplined users to retreat back to familiar, proven search methods.

While openness and conscientiousness lacked a direct line to usage, they were still linked to the technology through indirect pathways. All three personality traits showed a strong positive association with social image. Extraverts, highly open individuals, and highly conscientious people all tended to believe that mastering an advanced digital tool would make them look tech-savvy and elevate their status among their immediate peers.

This elevated social image then acted as a behavioral bridge. When users felt the chatbot boosted their reputation, their confidence in using the tool grew. This increased computer self-efficacy was strongly linked to how often they eventually logged into the program to complete practical tasks.

The link between social prestige and technical confidence relies heavily on how groups communicate about new software. When peers advocate for a specific tool, that verbal encouragement acts as a powerful source of motivation. Hearing a friend or colleague praise a program can reduce a user’s initial anxiety and make the learning curve feel much less intimidating.

Seeing other people succeed also provides a form of vicarious experience. If an individual watches a peer use a chatbot to complete a task quickly, they are much more likely to believe they can achieve the exact same result. This environmental feedback loops back into their own self-evaluation, steadily increasing their overall technical confidence.

The researchers noted this process forms a sequential chain of associations. A user’s personality influences their desire to look good in front of others. That desire for social standing correlates with a willingness to practice and feel competent. Ultimately, that tested competence translates into consistent usage.

Some personality traits had distinct limitations in building technical confidence. Extraverts, for example, did not show a direct increase in computer self-efficacy. Since extraverted people normally rely on facial expressions, vocal tone, and physical gestures to communicate successfully, a strictly text-based interface might limit their natural confidence. Highly open people also did not show a direct boost in technical confidence, possibly because their strengths lie in creative exploration rather than mechanical software mastery.

Conscientious individuals, conversely, displayed a robust relationship with computer self-efficacy. Their structured approach to learning likely helped them understand the software’s advanced features quickly. Once these individuals felt entirely capable, their usage of the application increased at a steady rate.

Software developers and technology marketers often study these behavioral pathways to refine their products. By understanding how different personality types respond to a new interface, companies can adjust user training modules to reduce early anxiety and build technical confidence. Future public campaigns might focus heavily on the social prestige of mastering the tool to attract status-driven consumers.

The dataset carries some notable limitations that future studies will need to address. The survey relied entirely on a cross-sectional design, meaning the researchers observed behavior at a single point in time. Because of this structural setup, the researchers could only identify mathematical associations between variables rather than prove that one factor directly caused another.

Relying on self-reported data also introduces subjective bias. Participants might overestimate or underestimate how much time they actually spend chatting with the program on an average day. Future investigations could pair software tracking metrics with psychological profiles to gather completely objective usage statistics.

The participant pool was relatively narrow, focusing mostly on young respondents living in China who had already adopted the technology. Including a broader range of ages, geographic locations, and language groups will help determine if these behavioral trends hold true globally. Comparing active users directly against people who refuse to use the technology might also provide a deeper understanding of digital reluctance.

The study, “Associations Between Personality Traits and ChatGPT Usage: The Dual Mediating Roles of Social Image and Computer Self-Efficacy,” was authored by Tingjun Deng, Dake Wang, Jiaojiao Ma, Tian Wang, Benqian Li, Talib Hussain, Yongjie Yue, and Pengcheng Wang.