🔄 The UFH AI Forum: A Catalyst for Change

The Artificial Intelligence for Multilingual Education Forum at the University of Fort Hare (UFH) recently underscored a pressing reality in South African higher education: the dominance of English as the primary language of instruction is creating significant barriers for many students. Hosted by UFH's Department of Business Innovation and Entrepreneurship in collaboration with the South African Centre for Digital Language Resources (SADiLaR), the event drew educators, researchers, technologists, and policymakers to discuss how artificial intelligence (AI)—specifically tools like real-time translation, speech recognition, and text-to-speech—can foster truly inclusive learning environments across South Africa's 11 official languages.

Dr. Vusumzi Funda, a key convener and researcher at UFH, emphasized the forum's role as more than dialogue—it was a 'critical call to action.' Participants highlighted that without swift adaptation to AI-enhanced multilingual strategies, millions of students risk being left behind in an increasingly digital academic landscape. This gathering, held in early 2026, aligned with broader national efforts to decolonize education and leverage technology for equity.

University of Fort Hare's Legacy and Vision

Established in 1916, the University of Fort Hare (UFH) holds a storied place in South Africa's educational history as the alma mater of icons like Nelson Mandela and Robert Sobukwe. Located in the Eastern Cape, UFH serves a predominantly rural and underserved student body, making it an ideal venue for addressing linguistic inclusion. The institution has pioneered AI applications, including an AI-powered chatbot that enhances student engagement and administrative efficiency.

UFH's commitment to multilingualism stems from its context: over 70% of its students speak isiXhosa or other African languages as their home tongue. Professor Roxanne Piderit, an associate professor at UFH, contributed insights on breaking language barriers through AI, praising the forum's thoughtful approach. This aligns with UFH's broader strategy to integrate innovation with cultural relevance, positioning it as a leader in AI-driven educational transformation.

Language Barriers in South African Higher Education

South Africa's higher education system grapples with profound linguistic challenges. Despite constitutional recognition of 11 official languages, English remains the medium of instruction in nearly all universities, affecting comprehension and retention. Studies indicate that students whose home language is not English face up to 30% higher dropout risks in their first year, contributing to overall first-year attrition rates of 50-60%.

For instance, at historically disadvantaged institutions like UFH, rural students often enter with strong foundational knowledge in subjects but struggle with English terminology in science, engineering, and law. This mismatch exacerbates inequities: black South African students, who comprise 80% of enrollees, experience proficiency gaps that hinder academic success and graduation rates hovering around 15-20% nationally.

• High first-year dropout: 35-52% linked partly to language.
• Low throughput: Only 6% graduate in minimum time.
• Rural-urban divide: isiXhosa, Sesotho speakers underrepresented in advanced programs.

AI as a Bridge to Inclusivity

Artificial intelligence offers transformative potential for multilingual education. Natural language processing (NLP) tools can provide instant translations, adaptive subtitles, and voice-enabled learning in African languages. SADiLaR, a national hub for digital language resources, plays a pivotal role by curating datasets for isiZulu, isiXhosa, Sesotho, and others, enabling AI models to 'understand' low-resource languages.

The Masakhane initiative, a community-driven effort, has developed open-source NLP models for over 50 African languages, empowering universities to create customized tools. At UFH, experiments with AI chatbots demonstrate improved student support, reducing administrative queries by 40% and boosting engagement among non-English speakers.

Photo by Zulfugar Karimov on Unsplash

Key Discussions and Expert Insights

The forum featured presentations on practical AI applications. Ntsundeni Mapatagane shared experiences from the event, highlighting partnerships for AI in higher education. Dr. Funda advocated for 'AI sovereignty' in language data, warning against reliance on foreign models biased toward high-resource languages like English.

Stakeholders debated ethical AI use: ensuring data privacy under POPIA (Protection of Personal Information Act) while scaling tools. Solutions included hybrid human-AI tutoring systems, where AI handles translation and instructors focus on cultural nuances.

Case Studies from South African Universities

UCT's Multilingual Education Strategy introduces isiXhosa glossaries for engineering, piloted successfully with 25% comprehension gains. Stellenbosch University employs AI transcription for lectures in Afrikaans and English, extending to African languages via SADiLaR resources.

At UKZN, workshops explore AI for multilingual pedagogies, with automated tools aiding transcription in Zulu and Xhosa. These cases show AI not replacing teachers but augmenting them, with early results indicating 15-20% retention improvements.

Challenges and Proposed Solutions

• Data Scarcity: African languages lack training data; solution: crowd-sourced corpora via SADiLaR.
• Infrastructure Gaps: Rural connectivity; hybrid offline-online AI apps.
• Teacher Training: Upskill via forums like UFH's; integrate AI literacy in curricula.
• Equity Risks: AI bias; ethical guidelines from national frameworks.

Forum outcomes include calls for a national AI-in-Education policy, prioritizing multilingual tools and funding for low-resource languages.

Stakeholder Perspectives and Impacts

Educators like Prof. Piderit stress cultural preservation: 'Language is identity.' Students report AI tools reduce alienation, with one UFH pilot showing doubled participation in discussions. Policymakers, via DHET (Department of Higher Education and Training), eye scaling these innovations amid 2026 enrollment pressures exceeding capacity.

Broader implications: improved employability, as multilingual graduates better serve diverse workplaces. Economically, addressing dropouts could add R100 billion to GDP via higher skilled workforce.

Photo by Hennie Stander on Unsplash

Future Outlook and Actionable Insights

Looking ahead, UFH plans AI labs for language tech development, partnering with Masakhane and SADiLaR. Recommendations for SA universities:

1. Audit language policies for AI integration.
2. Pilot text-to-speech in core modules.
3. Collaborate on shared datasets.
4. Monitor equity via annual audits.

For educators: start with free SADiLaR tools SADiLaR resources hub. Students: leverage apps like Google Translate customized for African languages. This forum marks a pivotal step toward an inclusive, AI-empowered higher education system.

Pathways Forward for South African Academia

In conclusion, the UFH AI Forum illuminates a viable path: harnessing AI to honor linguistic diversity while advancing equity. As South Africa navigates 2026's challenges—rising enrollments, funding strains—multilingual AI stands as a beacon for sustainable transformation.