Open Science infrastructures have greatly enhanced
global knowledge sharing, but the rise of AI presents new threats to the
integrity of scholarly communications. The rapid increase in AI-generated
content obscures the distinction between genuine research and sophisticated
fabrication, while aggressive data harvesting practices exploit open
repositories at scale.
In this session, we share practical experiences from
managing Zenodo, a large-scale open science infrastructure, focusing on how we
balance openness with security in the age of AI. We discuss technical solutions
such as advanced machine-learning classifiers to detect AI-generated spam,
governance strategies, and community-driven moderation policies that help
maintain the credibility and quality of open science systems.
Download the
presentation slides from Zenodo.