✅ **1. Creative Boost for the Less Inspired**
**Source:** UCL, University of Exeter — *Science Advances*, July 2024
**Paper Title:** *AI-enhanced creativity in story writing*
🔗 DOI: [10.1126/sciadv.adp3562](
https://doi.org/10.1126/sciadv.adp3562)
**Method:**
300 participants were asked to write stories. One group received creative suggestions from ChatGPT (either 1 or 5 prompts); the others received none.
**Results:**
Participants who received 5 ideas from ChatGPT produced stories rated as more original (+8%) and more useful (+9%).
The greatest improvement was observed in those with initially low creativity.
AI acted as a **creative catalyst**, not a substitute.
---
✅ **2. Enhanced Creativity in AI-Supported Teams**
**Source:** Rahman et al. — 2025, ResearchGate
**Title:** *Harnessing generative AI for collaborative creativity*
🔗 Link to study
**Method:**
University students performed creativity tasks (AUT - Alternative Uses Task) with or without AI (e.g., ChatGPT, Midjourney).
**Results:**
AI-supported teams outperformed the control group in fluency, originality, flexibility, and elaboration.
They also found the work more engaging and less stressful.
AI was perceived as a **group facilitator**, not a threat.
---
✅ **3. AI in Creative Educational Processes**
**Source:** *International Journal of Educational Technology in Higher Education*, May 2025
**Title:** *Collaborative digital storytelling enhanced by AI tools*
🔗 Read the full study
**Method:**
Student groups created digital stories with or without AI (ChatGPT, Runway, Midjourney). Assessed on problem-solving and creativity.
**Results:**
AI-supported groups showed significant improvements in:
* **Collaborative creativity** (TCP)
* **Complex problem-solving**
* **Efficiency and satisfaction** in creative work
Also noted: a stronger sense of inclusion in the creative process.
---
✅ **4. AI Comparable to Small Human Teams in Idea Generation**
**Source:** Cambridge Judge Business School, December 2024
**Title:** *Large Language Models as Creative Agents*
🔗 Official site
**Method:**
ChatGPT was given 13 creative tasks, each repeated 10 times. Outputs were compared to those of human teams.
**Results:**
AI models generated ideas comparable to those of an average human group (8–10 people).
In many cases, AI contributed **25–35%** of the top-rated ideas.
Conclusion: AIs are **competitive creative companions**, not just tools.
---
✅ **5. Artistic Value Perception: AI Holds Up**
**Source:** Tigre Moura — *The Journal of Creative Behavior*, 2023
**Title:** *Perceived Creativity of AI-Generated Works*
🔗 DOI: [10.1002/jocb.600](
https://doi.org/10.1002/jocb.600)
**Method:**
Public evaluation of poems, visual artworks, and music created by AI vs. humans.
**Results:**
Most participants couldn’t tell whether a work was human- or AI-generated.
AI outputs scored well in **originality, beauty, and emotional depth**.
No stigmatization: AI-generated art was judged **authentic**.