The amendments to the EU AI Act have mandated organisations globally to educate their employees on AI literacy. Wits CISO Galeboe Mogotsi discusses how higher education institutions are equipping themselves to meet the demands of this new regulatory era.
The new EU AI Act, coming into full effect in August 2026, is set to affect local companies and IT leaders overseeing cross-border operations, with AI literacy for employees being a cornerstone feature of the Act. This will have implications for both businesses and higher education organisations.
From a higher education perspective, Galeboe Mogotsi and his team are relying on partnerships for better connectivity and training opportunities.
“We’re working closely with local telecommunications partners and community stakeholders to expand broadband access and digital skills training,” he notes. “This includes initiatives like low-cost connectivity projects in underserved regions and community-based digital literacy workshops. By doing so, we aim to create a more level playing field so that all segments of the population can benefit from AI-related opportunities.”
According to Galeboe, Africa is poised for rapid economic growth through AI-driven solutions, but the pace of innovation can sometimes clash with ethical considerations.
“We can build internal AI ethics committees that bring together technologists, legal experts, ethicists, and community representatives,” he suggests. “This committee will evaluate AI proposals not just for their market potential but also for their alignment with core principles like transparency, fairness, and respect for local cultural norms.”
Galeboe went on to say that a responsible approach to AI could be a competitive advantage, so Wits has embedded ethical checkpoints into every stage of the AI development lifecycle.
“For instance, we can regularly consult with local communities and industry-specific partners to understand the cultural nuances that might affect data gathering or the deployment of AI-based decisions. By proactively tailoring our innovations to these cultural and ethical frameworks, we aim to drive both rapid growth and long-term trust in our AI solutions,” Galeboe adds.
Positioning African talent for success
“Internally, we must launch an ongoing training programme that covers everything from the fundamentals of machine learning to the legal and ethical considerations posed by AI. These sessions can be tied to different job roles, technical teams receive advanced technical training, while non-technical teams learn about AI risk management, oversight, and compliance,” Galeboe explains.
For Galeboe, getting the fundamentals of machine learning to the legal and ethical considerations posed by AI is a good start. “Externally, we can partner with all universities and tech hubs to sponsor AI bootcamps, hackathons and research collaborations. This ensures that emerging African talent gains both theoretical and hands-on experience with AI technologies.”
Algorithmic bias, which the EU AI Act specifically addresses by mandating that all providers of high-risk AI systems evaluate and mitigate biases in their training, testing and validation datasets, remains a concern for CIOs.
According to Galeboe, algorithmic bias can amplify existing social disparities, and in regions with historical inequalities, this risk is even more pronounced. “To mitigate it, we must adopt a ‘bias by design’ approach; every AI model must undergo rigorous bias audits before deployment,” he suggests.
Key provisions of the EU AI Act came into effect on 2 February 2025, with the next phase of implementation scheduled for 2 August 2025. The European Commission is expected to publish comprehensive guidelines by 2 February 2026, providing further clarity for organisations navigating compliance.
