Can generative artificial intelligence (AI) tools that create text, images and other content truly enhance employee creativity? A new paper published in the Journal of Applied Psychology and co-authored by Jing Zhou, Rice University’s Mary Gibbs Jones Professor of Management, finds that generative AI doesn’t enhance creativity equally for everyone. Instead, its impact depends on how well people manage their own thinking while using it.
To unlock generative AI’s creative potential, employees must pair it with “metacognitive strategies” — mental habits like planning tasks, tracking progress and adjusting tactics. To test how metacognition affects the creative benefits of generative AI, the authors ran a field experiment with employees at an actual organization — one of the first studies to do so. Using one of the most rigorous methods in organizational research, the study offers rare causal evidence of AI’s impact on workplace creativity.
The field experiment took place at a Chinese technology consulting firm, where employees are expected to develop custom solutions for diverse clients. The team worked with 250 nonmanagerial employees at the firm who were randomly assigned to either a treatment group or a control group. Treatment group participants received access to ChatGPT accounts along with usage examples, while those in the control group continued working without help from AI.
Both groups completed creative problem-solving tasks during the work week. At the end, participants in the treatment group took a survey measuring their use of metacognitive strategies, including how well they planned, monitored and adapted their methods. For example, they would rate themselves on statements like: “While working toward my goal, I kept track of how effective my approach was.”
Employee supervisors, who were unaware of the study’s design or purpose, later rated the creativity of participant outputs. To supplement these evaluations, external reviewers independently assessed employee responses to a specific written prompt about protecting privacy in the digital workplace. Responses were scored on novelty and usefulness.
The experiment showed that employees who worked with generative AI produced more creative ideas than those who did not — but only when they applied metacognitive strategies.
“What we found is that AI tools aren’t a creativity machine on their own,” Zhou said. “They can be a powerful partner. But for workers to truly benefit, they need to reflect on their thinking and adjust their approach in real time.”
The value of AI depends less on what it can do and more on how people use it, she explained.
“Tools like ChatGPT have real potential to expand what we call ‘cognitive job resources’ — that is, things like access to relevant information, the ability to switch between tasks and moments of mental rest,” Zhou said. “But this potential is only unlocked when people use it strategically.”
The study has major implications for both employees and companies who are eager to integrate AI into their workflows. If AI-driven creativity depends on how people think, then companies must invest not just in new tools but in the mental habits that make those tools work. And for employees, the study suggests using AI effectively isn’t just about mastering the technology — it’s about sharpening how you plan, adapt and reflect as you work.
“How and for whom using generative AI affects creativity: A field experiment” was published in the Journal of Applied Psychology by Zhou, Shuhua Sun (Tulane University), Zhuyi Angelina Li (Renmin University of China), Maw-Der Foo (Nanyang Technological University) and Jackson G. Lu (Massachusetts Institute of Technology).
This article was originally published in Rice Business Wisdom and was lightly edited for Rice News.