Among “Global Faculty Week 2025” lectures’ topics – the impact of AI on public administration and higher education

Artificial Intelligence (AI) has seeped into almost every sphere of our life, from education systems to the foundations of democratic governance. During Global Faculty Week 2025, academics from eight countries convened from May 5 to 9 in KTU Faculty of Social Sciences, Arts and Humanities to discuss about the growing role of AI in various fields, including public administration and higher education.

The week’s discussions highlighted that while AI offers great potential, it also brings significant social and ethical challenges – especially during digitalisation and in countries where low-resource languages are widely used.

AI raises questions of digital inequality

AI integration in South Africa’s higher education system threatens to deepen linguistic inequalities. While South Africa officially recognizes 12 African languages, English is still the main language used in universities and government offices due to the country’s complex colonial past.

Lerato Sono, a PhD candidate and lecturer at the University of Pretoria noted, that AI primarily “speaks” English, and this linguistic bias marginalizes students for whom English is not a first language.

These students often lack the proficiency needed to effectively use AI tools or critically assess the information generated by them. The result? A modern version of discrimination, where language becomes a barrier to using technology effectively.

Can we trust chatbots?

“Who do we hold accountable when a chatbot gives incorrect or misleading information?” asked Dr Stellah Lubinga, lecturer at the University of Pretoria. As South African government agencies begin using AI-powered chatbots on platforms like Facebook Messenger, new dilemmas are surfacing.

Her research investigates how unclear and confusing AI decision-making can be. Without clarity on how chatbots generate responses – or who bears responsibility for them – citizens may begin to lose trust in the very institutions meant to serve them.

Transparency, accountability, and digital literacy aren’t just popular buzzwords anymore –– they are essential for a healthy democracy in the age of AI.

Machine Learning in Public Service

If the South African case studies highlighted AI risks, Dr Cătălin Vrabie from Romania brought a more optimistic perspective. His research into using machine learning for image analysis in citizen petitions showed how AI can enhance – not hinder – public service.

Vrabie’s model uses image classification and object detection algorithms to analyse visual evidence submitted by citizens. Whether it’s a pothole, a broken streetlight, or illegal dumping, the AI model helps public officials respond faster and more effectively.

It’s a vision of e-government 3.0, where technology doesn’t just digitise bureaucracy but actively improves its responsiveness and reach.