Subscribe
About

Can South Africa power its AI revolution?

AI promises economic transformation, but demands vast energy resources, exemplifying the classic tension between innovation and failing infrastructure.
Rennie Naidoo
By Rennie Naidoo, Professor in Information Systems and Research Director at the Wits School of Business Sciences
Johannesburg, 13 Mar 2025
Rennie Naidoo, professor in Information Systems at the Wits School of Business Sciences.
Rennie Naidoo, professor in Information Systems at the Wits School of Business Sciences.

Artificial intelligence (AI) is transforming society at an unprecedented pace. Yet, it comes with a major challenge − AI is extremely energy-hungry, and South Africa’s power grid has been under immense strain for many years.

As AI integration accelerates rapidly across various industry sectors, a pressing question arises: Can South Africa fully embrace the AI revolution without exacerbating its ongoing energy crisis?

AI models require vast amounts of electricity − not only during training but also for continuous real-time operations. In fact, some estimates suggest that training a single model consumes as much power as thousands of households over several months. This growing demand makes a stable and reliable power supply just as critical as energy availability.

For South Africa, this presents a significant challenge. Load-shedding remains a persistent reality, and as AI adoption grows, it will further strain an already fragile grid. Without a stable and reliable power supply and sufficient energy availability, AI growth will remain crippled by Eskom’s ongoing failures.

A country that neglects basic infrastructure while chasing futuristic ambitions is not advancing − it is merely distracting itself from reality.

The real challenge is not just whether South Africa can embrace AI − it is whether it can do so without deepening its energy crisis. If Eskom fails to keep up, South Africa’s AI ambitions will be left in the dark − both figuratively and literally.

The paradox of AI and energy scarcity

South Africa faces a paradox. AI promises economic transformation, yet it demands vast energy resources that Eskom struggles to provide. This exemplifies the classic tension between innovation and failing infrastructure.

On one hand, AI-driven advancements can improve energy-efficiency, optimise grid management, and even help Eskom forecast and mitigate outages.

On the other hand, the very technology that could help solve these challenges also risks overwhelming an already overburdened power system.

This is the AI-energy paradox. Progress in AI depends on a reliable and scalable power supply, yet solving the energy and economic crisis may require the very AI innovations that an unstable grid cannot support.

AI dreams versus reality

South Africa’s leaders speak of a high-tech future, yet they fail to deliver even the basics. How can we talk about AI-driven progress when electricity is unreliable and roads are crumbling?

AI requires more than just ambition − it depends on stable electricity, high-speed connectivity and reliable infrastructure. Without these foundations, AI progress will remain out of reach.

Just as a high-performance car cannot drive smoothly on pothole-ridden roads, AI-driven smart cities cannot function on failing infrastructure. Power outages, deteriorating roads and mismanaged resources will only stall technological progress.

Ambitious AI innovations will fail under the weight of load-shedding, collapsing infrastructure and political mismanagement. No matter how promising the vision, it cannot succeed without the fundamentals in place.

Innovations are not built on ambition alone − it also requires talent, discipline and sound decision-making to create the right conditions for success.

In rugby, you do not win the World Cup by simply talking about greatness − you win by building a skilled team, executing a disciplined game plan and making smart, tactical decisions under pressure.

Innovation works the same way. No amount of ambition can compensate for a lack of competence.

True innovation depends on a skilled workforce, a strong work ethic and a culture that encourages calculated risk-taking − areas where South Africa continues to struggle.

Results over rhetoric

South Africa must choose competent leaders who prioritise results over rhetoric, infrastructure over inefficiency and accountability over excuses. The world will not pause while South Africa sorts out its fundamental problems.

A country that neglects basic infrastructure while chasing futuristic ambitions is not advancing − it is merely distracting itself from reality.

The laws of economics and reality do not bend for political convenience. A nation that cannot provide stable electricity, maintain functional roads, or appoint competent leaders should not expect to lead in AI.

AI does not operate in a vacuum. Without electricity, digital infrastructure and top talent, it remains a hollow promise. Rather than making lofty declarations about a high-tech future, elected leaders must first build and maintain the fundamentals needed to sustain it.

Without these essentials, AI will be just another buzzword, drowning in a sea of power cuts, bureaucratic inefficiency and failing institutions.

South Africa’s widespread mismanagement of scarce resources is not just an inconvenience − it is an active barrier to progress. Load-shedding cripples data centres, neglected roads weaken supply chains and political patronage kills ambition − driving away expertise and leaving innovation to stagnate.

Instead of leading in AI, South Africa risks becoming a spectator − watching from the sidelines as global competitors surge ahead.

Share