Four studies spanning short-story writing, circular-economy problem-solving, humour captioning and collaborative storytelling found that AI assistance raised the quality of individual creative output while narrowing the range of ideas produced across groups. Gains were largest for people with lower starting creativity, yet across all studies the collective pool of ideas became more similar when AI was used. The point in a workflow where AI entered mattered: applying it during idea generation reduced variety, whereas restricting it to idea selection kept diversity closer to levels seen in human-only work. Léonard Boussioux, Anil Doshi, Oliver Hauser and Kartik Hosanagar reported these findings in MIT Sloan Management Review on 9 July 2026, in an article examining how AI-assisted creativity affects organisations’ capacity for varied, breakthrough thinking over time.
Source: MIT Sloan Management Review, 2026-07-09.