Source: Sun et al. (2025) — “How and for Whom Using Generative AI Affects Creativity: A Field Experiment,” Journal of Applied Psychology
AI creativity tools are easy to oversell.
Give everyone generative AI, and suddenly the workplace becomes more creative. Ideas everywhere. Innovation blooming. Someone from strategy says “step change” and everyone pretends not to flinch.
This paper asks a better question: how and for whom does generative AI affect creativity?
The “for whom” part matters.
The authors focus on metacognitive strategies, which is a fancy way of saying people who monitor and regulate their own thinking. They do not just produce ideas. They think about how they are producing ideas.
That skill changes how AI is used.
If someone has strong metacognitive habits, AI becomes a resource. They can use it to expand options, test directions, challenge assumptions, and then choose. They stay in charge of the thinking process.
If someone lacks those habits, AI can become a shortcut that feels productive but may not lead to better creativity. The person receives outputs, but does not necessarily know how to interrogate them.
This is exactly why AI training should not only teach prompting.
Prompting is useful, yes. But the deeper skill is reflective thinking. What am I trying to solve? Why is this suggestion good or bad? What assumption is hiding here? What would make this idea actually useful?
Without that, AI can produce a lot of options and still leave you with no taste.
Which is basically a buffet problem, but for ideas.
My takeaway: generative AI can support creativity, but it rewards people who know how to manage their own thinking. The future creative advantage may not be having AI. Everyone has AI. It may be knowing how to think with it without letting it drive.