Communicating science in the age of GenAI: Can generative AI support the writing of better science communication products?
Journal article › Research › Peer reviewed
Publication data
| By | Barel-Ben David Yael, Roy Yosef, Elad Segev |
| Original language | English |
| Published in | Science Communication |
| Pages | 27 |
| Editor (Publisher) | SAGE Publications |
| ISSN | 1075-5470 |
| DOI/Link | https://doi.org/10.1177/10755470251411176 |
| Publication status | Published advanced online – 02.2026 |
Effective science communication (SciComm) is crucial, but training scalability remains challenging. We explored whether generative AI (GenAI) could provide feedback to enhance SciComm strategies. In an online iterative distillation exercise, SciComm trainees (N = 78) condensed their research. An experimental group (n = 41) received jargon-oriented and GenAI feedback; controls (n = 37) received only jargon feedback. Participants preferred revised texts, with slightly higher preference in the GenAI group. SciComm-based rubric assessment revealed GenAI-supported texts significantly improved in SciComm strategies, particularly connecting science to everyday life and narrative use. Findings highlight GenAI’s potential to enhance SciComm content and scalable feedback, supporting its careful integration into training.
