By: Steven Santini - Vice President for Secure Power, SSA at Schneider Electric

In 1969, the world watched in awe as Neil Armstrong took “one small step for (a) man, one giant leap for mankind”, a defining moment that represented how far we had come as a species. It represented scientific leadership, industrial capability and ultimately, a glimpse into our future.

Fast-forward to 2026, and the manned crew of Artemis II has now paved the way for future human surface missions. However, it also a year that has given way to another “moon race”, if you will, the ability to secure, generate and distribute enough electricity to support AI infrastructure at scale.

Indeed, for much of the last decade, the global technology conversations have centred on compute power and semiconductor availability; this has changed dramatically.

According to the International Energy Agency (IEA), global data centre electricity consumption will, by 2030, reach nearly 945 TWh (under baseline forecasts), which represents almost 3% of all global electricity consumption

Furthermore, the IEA notes that electricity consumption from accelerated AI servers is expected to grow by roughly 30% annually this decade. To place this into perspective, a single hyperscale AI data centre can consume as much electricity as a small city.

And this is where the proverbial penny drops; most existing power grids, built decades ago, were simply not designed for this level of concentrated, always-on demand. 

Ready, set

This above had therefore led to major behavioural change or “race” by major AI infrastructure players.

Rather than relying solely on national utilities, many prominent hyperscale operators are now pursuing direct control over energy supply through long-term power agreements, private generation partnerships and dedicated infrastructure investments.

Some are even aligning with emerging small modular nuclear reactor programmes, while others are investing in large-scale renewable generation paired with battery storage. Several have begun securing their own energy assets outright in order to guarantee future supply capacity.

Substantiating the above, is a recent IEA Energy Supply for AI report which highlights that  global electricity generation dedicated to data centres could exceed 1 000 TWh by 2030, with renewables, natural gas and nuclear expected to carry most of the expansion burden.

Africa is called

This bring is back to why the African continent has now become part of this “moon race”.

For years, discussions about Africa’s digital future focused primarily on connectivity, mobile adoption and cloud migration. Now, the continent is being evaluated as a potential frontier for next-generation AI infrastructure.

Africa offers available land, expanding fibre connectivity, growing investment openness and, critically, untapped energy development potential. Several markets are accelerating renewable energy programmes, independent power projects and regional interconnection initiatives that could support future digital infrastructure growth.

That said, the continent is now without its challenges; grid stability, transmission limitations, financing constraints and policy certainty remain significant hurdles. However, what the above does signify is that Africa is increasingly part of a global strategic discussion around where future AI infrastructure can realistically scale.

A race bigger than technology

What makes the race for power also important is that it exposes a deeper truth about AI. Often perceived as abstract software-driven operations living the cloud, it is anything but.

Today, every AI query, training model, inference engine and generative workload depends on real-world infrastructure: substations, cooling systems, transformers, cables, backup power, battery storage and generation capacity.

The emerging AI race is therefore no longer simply about software innovation. It is becoming a contest of industrial capability, energy resilience and infrastructure execution.

Those nations, regions, and organisations that master the integration of power generation, modern grids, energy storage, and digital infrastructure will be the ones breaking the tape at the finish line and then start reaching for the moon.