<p>Sequencing Human-Artificial Intelligence Creative Writing Collaboration Affects Perceived Authorship</p>

As generative AI tools enter domains traditionally considered uniquely human, questions arise about their impact on creative processes. This study tests how the sequencing of human and AI contributions in creative writing affects perceived creativity, authorship, and satisfaction. In a between-subjects experiment (n = 84), participants either wrote a story and then used generatestory.io to improve it (AI-enhanced) or revised a draft generated by the tool (AI-generated). Results show that participants reported similar satisfaction in both conditions. However, authorship feelings differed strongly after the first step (human writing first vs. receiving an AI draft) and became much more similar after people revised the story. Mediation analyses suggest that early perceived authorship helps explain later satisfaction, whereas perceived creativity does not. Overall, the findings imply that who contributes first mainly shapes early feelings of ownership and that hands-on revision can restore perceived authorship, which matters for designing co-creative writing tools that keep users engaged and satisfied without undermining their sense of being the author.