Open-Source Generative Artificial Intelligence and Open Science: Global Developments and Challenges

Abstract

Background and Purpose: Open data and research outputs form the backbone of Open Generative Artificial Intelligence (OpenGenAI) models. Based on core principles of openness, transparency, reusability, and extensibility, these technologies are intrinsically tied to information sovereignty. This research aims to examine the fundamental principles of OpenGenAI, monitor its global development and implementation, and identify key challenges to ensure its application benefits society.

Methods: The methodology involves a literature review, as well as descriptive, explanatory, and predictive methods. Four levels of research are applied, and for exploratory analysis, several resources from governments, research organizations, and universities are consulted.

Results: Information search and retrieval are based on eight defined categories, and based on the findings obtained, a set of challenges regarding this technology is identified.

Conclusion: A general conclusion is that when OpenGenAI is managed by governments, research organizations, and universities, it will address inequalities in access. This will enable societies to benefit from responsible and democratic information products generated by the technology.