Mapping The Use of Generative Artificial Intelligence in Group Creative Processes: A Scoping Review

Creativity has long been understood as a spark that happens between people. When a team of designers, writers, or students gathers to solve a problem, the magic often lies not in any single mind, but in the friction and flow of their interaction. They build on each other's thoughts, challenge assumptions, and combine different viewpoints to create something new. For decades, technology has been a tool in this process, helping groups communicate or organize their work. But a new kind of technology has arrived that does more than just organize; it creates. Generative artificial intelligence can now write stories, draw images, and compose music on its own. This raises a profound question for anyone who works in a group: when a machine joins the circle of creators, does it help the group find better ideas, or does it change the way they think together?

A team of researchers from the Università Cattolica del Sacro Cuore in Italy set out to map exactly how this new technology is being used in group settings. They conducted a wide search of scientific studies to find real-world examples where at least two people worked together on a creative task while using generative artificial intelligence. They looked for evidence of how these tools were used, what kinds of work they helped with, and what happened to the group's dynamic. After reviewing thousands of records, they identified thirty-eight studies that fit their criteria. These studies covered a range of activities, from brainstorming sessions in classrooms to design workshops and professional innovation teams. The researchers found that while the technology is showing great promise, its role is complex, bringing both significant advantages and new, tricky challenges.

The most common setting for these experiments was education, where students used the tools for projects, followed closely by design and creative industries. The tasks varied, but the most frequent activity was brainstorming or generating new ideas. In these scenarios, the artificial intelligence was rarely just a passive tool sitting in the corner. Instead, it most often acted as a co-creator, an active partner that contributed to the flow of the work. The researchers observed that the technology was usually brought in during the early stages of a project, when the group was trying to generate as many possibilities as possible. It was less common to see the technology used for the later stages of judging, refining, or finalizing a creative product.

When the groups used these tools, the results were generally positive, but with a clear pattern. The most frequent benefit was a boost in the quantity and variety of ideas. The artificial intelligence could suggest unexpected connections or offer visual prompts that helped the group break out of mental ruts. It also made the work faster, handling the tedious parts of drafting or organizing information so the humans could focus on higher-level thinking. In many cases, the groups reported feeling more engaged and motivated, finding the process more enjoyable. The technology seemed to act as a shared focus point, a digital artifact that the whole team could look at and discuss, which helped them coordinate their efforts and build a shared understanding of the problem.

However, the study also highlighted significant tensions that arise when a machine becomes a creative partner. The most common worry was that people might rely too heavily on the computer, letting it do the thinking for them. This "cognitive offloading" meant that group members sometimes stopped questioning the ideas the machine produced or stopped generating their own original thoughts. There was a fear that the output would become too uniform, with everyone's work looking the same because they were all using the same type of tool. This could reduce the diversity of perspectives that makes group creativity so powerful.

Another major issue was the blur between human and machine ownership. When a group creates something with the help of an artificial intelligence, it becomes difficult to say who is responsible for the final result. Who owns the idea? Who gets the credit? The researchers found that this uncertainty created anxiety in some groups, particularly in professional settings where accountability matters. Additionally, the technology sometimes disrupted the natural rhythm of human interaction. In some cases, people stopped talking to each other and started talking only to the machine, or the machine's suggestions interrupted the flow of conversation in ways that felt intrusive.

The researchers concluded that the technology is not a magic solution that automatically makes groups more creative. Instead, it is a powerful new element in a complex system. The success of the collaboration depends less on the capabilities of the software and more on how the group chooses to use it. If a team uses the tool to expand their thinking while maintaining their own critical judgment and keeping the conversation flowing between people, the results are often excellent. But if they let the tool take over the thinking process, the benefits can quickly turn into drawbacks. The future of group creativity with artificial intelligence will likely depend on designing workflows that keep humans in charge of the direction, ensuring that the machine remains a partner in the process rather than the pilot.