AI Podcasts: Making Science Accessible
- Artificial intelligence can produce podcasts so realistic that listeners frequently enough cannot distinguish them from human-created content.
- Professor Philip Moons, of the University of Leuven in Belgium, led the research, which was published in the European Journal of Cardiovascular Nursing (EJCN).
- Moons presented the findings at the Association of Cardiovascular Nursing.
AI-Generated Podcasts Open New Doors to Science Communication
Updated June 14, 2025
Artificial intelligence can produce podcasts so realistic that listeners frequently enough cannot distinguish them from human-created content. A new study shows half the authors thought the podcasts were human.
Professor Philip Moons, of the University of Leuven in Belgium, led the research, which was published in the European Journal of Cardiovascular Nursing (EJCN). The team used Google NotebookLM, Google’s AI research assistant, to create podcasts explaining research recently published in the journal.
Moons presented the findings at the Association of Cardiovascular Nursing. The study marks the frist instance of AI generating podcasts about scientific papers.
Researchers are now exploring how the public perceives AI-generated science communication, including podcasts. They are also considering using AI podcasts to summarize scientific conference sessions.
“For instance, making a podcast on the content of sessions at the ESC conferences, as a summary for those who did not participate and want to get a blurb of it,” Moons said.
What’s next
The researchers have already created an AI podcast about their research paper. The podcast is linked in the published paper.