In a pivotal move for South Africa's technological future, Professor Benjamin Rosman, Director of the University of the Witwatersrand's Machine Intelligence and Neural Discovery (MIND) Institute, has been appointed to chair an independent expert panel tasked with rebuilding the nation's draft National Artificial Intelligence Policy. This development comes just weeks after the government withdrew the initial draft amid revelations of fabricated citations, underscoring the need for rigorous, evidence-based policymaking in artificial intelligence.

The appointment by Communications and Digital Technologies Minister Solly Malatsi highlights Wits University's growing leadership in AI research across Africa. Rosman, a professor in the School of Computer Science and Applied Mathematics, brings decades of expertise in machine learning, robotics, and AI ethics, positioning him ideally to guide this critical national effort.

Professor Benjamin Rosman's Trailblazing Career in AI

Benjamin Rosman is no stranger to pioneering AI advancements. Named one of Time magazine's 100 Most Influential People in AI in 2025, he has co-founded major African AI initiatives like the Deep Learning Indaba, fostering continent-wide collaboration. At Wits, his work spans foundational principles of intelligence in machines, humans, and animals, blending computer science with neuroscience, psychology, and ethics.

Rosman's research emphasizes algorithmic sovereignty for Africa—ensuring technologies reflect local contexts rather than imported biases. "I've spent much of my career thinking about how South Africa and Africa can participate meaningfully in AI's future," he noted, warning against becoming mere consumers of foreign systems optimized for other societies.

The MIND Institute: Wits' Hub for African AI Innovation

Central to Rosman's leadership is the MIND Institute, launched in late 2024 as an interdisciplinary powerhouse. Combining Wits' strengths in AI, machine learning, robotics, and autonomous systems with insights from philosophy, law, and humanities, MIND aims to unlock intelligence frontiers relevant to African challenges.

The institute drives postgraduate training, research networks, and practical applications, from healthcare diagnostics to sustainable agriculture. Google.org's US$1 million grant in 2025 bolstered its efforts, recognizing its potential to transform AI research on the continent. Under Rosman, MIND positions Wits as Africa's AI vanguard, directly informing national policy.

The Citation Scandal That Prompted Policy Overhaul

The catalyst for the panel was the April 10, 2026, release of the draft National AI Policy, which proposed ambitious institutions like a National AI Commission, AI Ethics Board, and AI Safety Institute. However, a News24 investigation exposed at least six fabricated citations out of 67 references—hallucinations from generative AI tools, including non-existent articles in real journals like AI & Society.

Minister Malatsi swiftly withdrew the document, calling the lapse "irresponsible" and implementing internal safeguards. Editors from affected journals confirmed the fakes, eroding credibility at a time when AI governance demands transparency. This incident spotlighted risks of unverified AI in policymaking, prompting a reset grounded in verifiable evidence.

Panel Composition: A Blend of Expertise

The seven-member panel unites top minds: alongside Rosman, Professor Vukosi Marivate (University of Pretoria, AI researcher), Dr Tshepo Feela (AI specialist), Dr Jabu Mtsweni (CSIR), Professor Alison Gillwald (Research ICT Africa), Advocate Lufuno Tshikalange (legal expert), and attorney Heather Irvine. Spanning academia, industry, civil society, and law, they ensure diverse perspectives.

• Research Focus: Marivate and Feela on technical AI advancements.
• Governance: Gillwald and Mtsweni on policy and digital infrastructure.
• Legal: Tshikalange and Irvine on ethical frameworks and compliance.

This multidisciplinary approach mirrors MIND's model, promising robust recommendations.

Photo by muhammad iqbal on Unsplash

Mandate: Balancing Protection, Innovation, and Capacity

The panel's charge is to craft a policy protecting citizens from AI harms—like bias in hiring or surveillance—while fostering economic growth and building national capabilities. Rosman stresses three pillars: harm mitigation, innovation enablement, and long-term skills development. "If we only regulate, we miss opportunities; if only innovate, risks deepen inequality," he explained.

Consultation will engage stakeholders from research, labor, industry, and civil society, tailoring to South Africa's realities: high unemployment, digital divides, and constitutional values like equity. For details on the process, see the Wits announcement.

Implications for South African Higher Education

This panel elevates universities' role in national AI strategy. Wits, through MIND, exemplifies how higher ed can drive policy via evidence-based research. Other institutions like UCT and UP contribute via panelists, fostering ecosystems for AI startups, training, and ethics labs.

Universities face opportunities: increased funding for AI programs, interdisciplinary centers, and global partnerships. Yet challenges persist—faculty shortages, infrastructure gaps, and brain drain. A strong policy could unlock grants, tax incentives for AI R&D, and job creation in data science and robotics.

AI Research Boom at South African Universities

South Africa's higher ed sector is AI-ready. Wits leads with MIND's focus on foundational intelligence; UCT excels in health AI; Stellenbosch in agritech applications. The Deep Learning Indaba, co-led by Rosman, trains thousands annually.

Stats show growth: AI publications tripled since 2020; enrollments in computer science up 40%. Policy success could integrate AI into curricula, addressing 30% youth unemployment via skills like machine learning engineering. For panel insights, read TechCentral's coverage here.

Challenges in African AI Governance

Africa lags global AI leaders—China, US, EU—with fragmented policies. SA's effort counters this, but hurdles include data sovereignty, ethical biases in imported models (e.g., facial recognition failing dark skin), and infrastructure deficits (only 40% broadband access).

Rosman warns of "algorithmic colonialism," urging local models trained on African data for languages like isiZulu. Panel must address risks: job displacement in call centers, privacy in welfare systems.

Opportunities: Jobs, Growth, and Global Standing

Optimism abounds. AI could add R500 billion to SA GDP by 2030, per McKinsey estimates adapted locally. Universities train talent for sectors like fintech (e.g., Capitec AI lending), mining autonomy, and precision farming.

• Job creation: 100,000+ in AI/data roles by 2030.
• Innovation hubs: Wits AI accelerators partnering startups.
• Equity: Policies mandating inclusive datasets.

Higher ed benefits via research grants, attracting talent like Rosman.

Photo by Vitaly Gariev on Unsplash

Stakeholder Perspectives and Consultations

Industry welcomes evidence-focus; Stafford Masie cautions against over-regulation stifling startups. Civil society pushes human rights integration. Labor unions eye reskilling funds.

Panel's consultative model ensures buy-in, drawing public input before resubmission—potentially late 2026.

Future Outlook for SA AI and Higher Education

Rosman's panel could position SA as Africa's AI leader, with universities central. Expect policy rollout by 2027, spawning institutes, incentives, and curricula reforms. Wits MIND exemplifies: from robotics labs to policy influence.

For academics eyeing AI careers, opportunities abound in governance, ethics, and application. South Africa's blend of talent, policy ambition, and urgency makes it fertile ground. As Rosman puts it, this is about building capabilities for an AI future that serves all South Africans.