GOVERNANCE OF GENERATIVE ARTIFICIAL INTELLIGENCE IN HIGHER EDUCATION INSTITUTIONS: AN APPROACH FROM DIGITAL SECURITY AND DATA PROTECTION

Abstract

The purpose of this research is to examine how generative artificial intelligence is managed in universities, focusing on digital security and data protection. This is done through an exhaustive review of the literature following the PRISMA protocol. To carry out this study, a thorough search was carried out in the Scopus and Web of Science databases, initially identifying 31 investigations. After eliminating duplicates and applying inclusion and exclusion criteria, 19 scientific articles were selected for further analysis. The findings show that generative AI governance is a new and complicated arena, combining technological, organizational, ethical, and regulatory factors. It was found that the implementation of these technologies in universities has exceeded the capacity of these institutions to create adequate regulations, which entails risks on digital security, data protection and the integrity of academic processes. Fragmentation was also noted in the scientific literature, with limited integration between governance, cybersecurity, and data management approaches, making it difficult to advance in the development of consistent institutional models.
Among the most important findings, it is highlighted that the most significant dangers are linked to the vulnerability of digital systems, the exposure of sensitive information, the lack of transparency in algorithms and the absence of digital skills in the academic community. It was also shown that artificial intelligence can act on two levels, being both a source of risk and a resource to improve information security. Therefore, generative AI governance needs a comprehensive approach that unites technological innovation with cybersecurity tactics, data protection policies, and digital skills training. 
In conclusion, the study indicates that higher education institutions must work on creating governance frameworks that are dynamic, adaptable and context-specific, allowing them to balance the adoption of emerging technologies with the safeguarding of information and user rights. Practical implications are also suggested that seek to strengthen institutional capacities, especially in regional contexts such as Latin America, Colombia and Barranquilla, where important challenges persist in the management of artificial intelligence.