Do AI tools undermine our sense of creativity? New study says yes

A new study has found that people tend to feel less creative when using artificial intelligence tools, even if they consider themselves generally creative. Published in The Journal of Creative Behavior, the research highlights how beliefs about personal creativity shift in AI-assisted contexts, and how these beliefs relate to creative activity and achievement.

As artificial intelligence tools become increasingly common, understanding their effects on how people see themselves is becoming more urgent. One area of concern is creativity. Some have predicted that AI will revolutionize creative work, while others worry it could make human creativity feel irrelevant. Despite this debate, little research has examined how people’s beliefs about their own creativity change when they’re using AI. This study was designed to explore that question and uncover how these self-perceptions relate to actual creative behavior and success.

“Let’s start with a brief premise: in order to achieve a creative result, such as writing a song, designing a logo, or even developing a new drug, it’s necessary to engage in creative activities,” said study authors Angela Faiella of the University of Bologna and Aleksandra Zielińska of the University of Wrocław.

“To begin these activities, we need a creative drive, i.e. creative self-belief: the motivation and confidence in our own abilities as creatives in that specific field. In other words, we act creatively if we feel motivated and capable of doing so. With artificial intelligence (especially generative AI), we asked ourselves: what happens to our creative self when we use AI tools to create? Do we still feel like creatives? Or do we feel replaced? In practice, we wanted to understand how our perception of our creative self changes when AI comes into play, compared to when we don’t use it.”

To address these questions, the researchers recruited 273 adults from the online platform Prolific in December 2023. All participants were native English speakers who had previously used AI tools. They completed an online survey that asked them about their creative self-beliefs, both in general and specifically when using AI. The survey also asked about their engagement in creative activities — like writing, visual art, or music — both with and without AI tools, and about any creative accomplishments they had achieved. Participants also answered questions about their personality traits and how much they trusted AI.

Free daily newsletter

To measure creative self-beliefs, the researchers used a well-established questionnaire that includes two main parts: creative confidence (how much people believe in their ability to be creative) and creative centrality (how important creativity is to their identity). For the AI-specific version, participants answered similar questions, but focused on how they felt about their creativity while using AI tools.

The results showed a significant gap between general and AI-specific creative self-beliefs. On average, participants rated themselves as more creative in general than when working with AI. This was a fairly large difference. While general and AI-specific self-beliefs were positively correlated, they only overlapped to a limited degree — suggesting they are related but distinct aspects of how people see themselves. People with low general creative self-beliefs were unlikely to feel creative with AI. But having high general creative self-beliefs did not guarantee feeling creative when using AI tools.

The researchers also examined how trust in AI influenced this relationship. People who had more trust in AI tools were more likely to feel creatively confident when using them, even after accounting for their general creative self-beliefs. However, trust didn’t completely close the gap between general and AI-specific beliefs. Even those who were confident and trusting still tended to feel less creative in AI settings.

One key question the study explored was whether AI-specific creative self-beliefs play a meaningful role in people’s real-world creativity. The researchers found that people who felt creative while using AI were more likely to engage in creative activities with AI. In turn, this creative engagement was linked to greater creative achievement. However, AI-specific self-beliefs were not directly related to achievement — only indirectly through activity. Strikingly, when the researchers controlled for other factors like general self-beliefs and personality, the direct link between AI-specific self-beliefs and creative achievement was actually negative.

This unexpected result suggests that people who strongly believe in their AI-specific creative abilities — independent of how creative they feel in general — might be relying too heavily on AI tools. The researchers speculate that this could reflect a kind of over-attachment to AI, where people begin to attribute creative success to the tools rather than their own efforts. In contrast, general creative self-beliefs were both directly and indirectly associated with creative achievement, whether or not AI was involved.

The findings also highlight how personality traits influence creative self-beliefs. General creative self-beliefs were most strongly associated with openness to experience — a trait linked to curiosity, imagination, and a willingness to explore new ideas. AI-specific creative self-beliefs, however, were more strongly tied to traits like extraversion and agreeableness, which are typically associated with social interaction. The researchers suggest this may be because people often interact with AI tools in ways that mimic conversations or collaboration, even if the AI isn’t a real social partner.

Another important finding was that people’s experience using AI in creative tasks mattered. Those who had more practice working with AI tools in creative contexts reported higher AI-specific self-beliefs. This suggests that self-perceptions can shift over time as people become more familiar with AI and gain confidence in using it to support their work.

“Our study showed that people feel less creative when using artificial intelligence for creative tasks compared to when they don’t use it,” Faiella and Zielińska told PsyPost. “AI can certainly help speed up the process and achieve creative results, but it doesn’t automatically make us feel more creative.”

“And this is a crucial point, because feeling creative means having confidence in our own ideas and feeling motivated to realize them, an essential drive to start and carry out creative activities. We believe AI shouldn’t be seen as a shortcut or a simple ‘response generator,’ but as a tool that can enrich and support our creativity without replacing us. If used mindfully, it can stimulate new questions and provide valuable input for finding original solutions.”

The study has some important limitations. Because the data were collected at a single point in time, the researchers cannot say for sure whether self-beliefs lead to creative activity or vice versa. It’s also unclear how lasting these effects are, or whether people’s attitudes change as they gain more experience with AI.

“The data were collected in 2023, at an early stage of widespread AI adoption: many people were still becoming familiar with these tools, and this may have influenced the scope and nature of the creative experiences analyzed,” the researchers noted.

Despite these limitations, the findings suggest that the way people feel about their own creativity in AI settings has real consequences. If people feel less creative when using AI — even if the output is impressive — they may become less motivated to engage in creative tasks, or may fail to recognize their own contributions. This could have broader implications for education, design, the arts, and other fields where creativity is valued.

“In the long term, we want to apply these findings in educational contexts, for example in university courses (which we are already doing), to promote mindful AI use,” Faiella and Zielińska said. “AI can be dangerous if not used in a proper way. However, we don’t think telling designers or creatives not to use AI is the solution. Rather, we want to show how to use it without diminishing individual creative confidence, encouraging co-creativity between humans and AI.”

“We hope our work sparks discussion about human–AI interaction. We believe that mindful use of AI in creative processes is essential. AI can be used in many ways, some more reflective than others, that can either help or hinder us. It can dampen our motivation to engage in creative tasks or profoundly transform how we approach them. Understanding these dynamics is essential to prevent AI from reducing our creative self-beliefs.”

The study, “Am I Still Creative? The Effect of Artificial Intelligence on Creative Self-Beliefs,” was authored by Angela Faiella, Aleksandra Zielińska, Maciej Karwowski, and Giovanni Emanuele Corazza.