The tail end of May brought with it a change of temperatures, the falling of leaves, and a fantastic discussion with Red Hat.

This meant of course, an abundance of insights and an afternoon of enlivening conversation.

The first point raised was that businesses across South Africa face a sizable challenge of embracing modernisation.

While that sounds like an easy task at first glance, it is anything but, because Red Hat explained that it must be done while striking a balance between rising technology costs and the need to build future-ready and compliant systems and applications.

The company noted that while the market for cloud computing services continues to grow, the focus is turning away from infrastructure migration. Instead, businesses are turning their attentions towards being more outcomes based. In practical terms, this means businesses are looking at how they can best use their infrastructure to transform operations and generate new value.

“The goal for local enterprises is to not only do more with less, but do it in such a way that it doesn’t compromise future growth,” stressed Dion Harvey, the Regional General Manager at Red Hat Sub-Saharan Africa.

“Modernisation is now not just about migrating legacy systems onto new platforms. It’s about creating environments where innovation can thrive and developers have the means to explore new project opportunities, as well as easily scale and adapt as their organisations evolve,” he elaborated.

That goal was front and centre at the recent Red Hat Summit 2026, the world’s premier enterprise open-source event. The annual event featured multiple new product releases and announcements, customer success stories and ecosystem partnerships that Red Hat’s vision for open source at a time when enterprise IT is becoming more complex and critical to businesses.

“The organisations that will have the most success in this environment are not necessarily going to be the ones that spend the most or the ones that move the fastest. They will be the ones that build on the right foundation,” said Matt Hicks, CEO of Red Hat, during the summit keynote speech. “Virtual machines, containers and AI agents are not competing with each other. They are converging. And the answer to this is not going to be one cloud or one vendor or one model; it will be the right platform with a broad ecosystem that supports it.”

What was particularly interesting is that the summit, which took place in Atlanta, Georgia, last month, spoke even more equally to South African businesses as it did to their international counterparts, Harvey noted. This suggests that the playing field has well and truly levelled out, at least in some respects. 

Red Hat AI and agentic systems: The next phase of the AI revolution

Admittedly, the large international conferences are often fascinating to attend. On the flip side, they can also feel like worlds away from a South African perspective.

‘’We are different. We have customers who are super innovative, but we have just as many customers who are more cautious in their approach,’’ Harvey pointed out. But this year, he related, something had changed.

‘’It truly felt like Red Hat was understanding the context, the practical context of what our customers here in South Africa, are going through today,’’ he added.

Harvey continued that AI is not the only thing on customers’ minds. ‘’In fact, it's probably number three or number four in terms of practical things to get done. There are far bigger challenges. Our customers are still trying to modernize, they are still trying to get to grips with how they leverage the cloud, they're still trying to get to grips with keeping the lights on, so to speak, in terms of modernizing what is still an increasingly disparate world that they operate in.’’

Localising the international AI trend

With that said, AI is still the juggernaught on the world stage, the elephant in the room, and the unmissable tech titan dominating headlines and deadlines.

So where exactly is the country when it comes to AI adoption?

According to the Red Hat team, South Africa remains on a steady AI adoption path as enterprises take their projects from experimentation to production. Central to that adoption are platforms that serve as a trusted foundation supporting any model or agent, running on any hardware accelerator across different cloud environments.

With that foundation, the company noted, local enterprises can develop, run and move workloads as they see fit, while transforming traditional business functions and workflows with AI-enabled applications.

In light of that, Red Hat has made significant advancements across its AI portfolio of flexible, cost-effective solutions, simplifying the development and deployment of agentic workflows and letting organisations scale AI across infrastructure. The release of Red Hat AI 3.4 marks the delivery of Model-as-a-Service (MaaS), which provides a single, governed interface for developers to access curated models, as well as the prompt management and an evaluation hub for assessing model and agent quality, accuracy and security.

Supporting the enterprise adoption of AI and autonomous agents is Red Hat’s AI Factory with NVIDIA software platform, engineered in collaboration with the global chipmaker and hardware manufacturer. As part of the latest updates to the platform, Red Hat AI Factory with NVIDIA has received enhanced security capabilities for continuously operating agents, which the company explained ''is further optimised to serve as a unified, scalable foundation for production-grade agentic AI.''

“Agentic AI is already being implemented in South Africa in key sectors such as financial services. The revolution is already taking place, and by refining platforms that democratise access to these systems and make them more adaptable to business needs, local enterprises can deploy agents of their own,” Harvey explained.

Linux, the optimal OS for AI agents across hybrid cloud

Critical to the growth of agentic AI is the ability to build and deploy agents. This means the choice of operating system (OS), and how it supports rapid innovation, is also imperative. “The shift towards autonomous AI workflows means developers need to look beyond standard Linux distributions and use one that embodies an agent-native software factory,” Harvey asserted.

To this end, Red Hat highlighted its Fedora Hummingbird Linux which is a freely available, container-native, rolling Linux release designed for agent-first builders. Fedora Hummingbird Linux supports agent-driven pulls for near-instant deployment across hybrid environments and streamlines the development process from proof-of-concept to scalable, supported production.

Combined with new developer tools from Red Hat that provide a standardised path for building and scaling agentic systems from local workstations to the hybrid cloud, including Red Hat Desktop and the enhanced Red Hat Advanced Developer Suite, agentic AI is within closer reach for South African businesses.

“Local businesses also need to keep the full lifecycle of their applications and infrastructure in mind. Value may be unlocked over a certain period and through different software iterations. This is why Red Hat also offers a continuous support path for systems with the Red Hat Enterprise Linux Long-Life Add-On. With capital spent on infrastructure, businesses need platforms that offer long-term returns,” Harvey continued.

The rise of the sovereign cloud

As enterprises modernise operations and adopt AI, organisations are also rethinking how digital infrastructure and data are managed. Africa, along with other continents, is becoming increasingly concerned with digital sovereignty, a topic that encompasses the ownership of digital infrastructure as well as the storage and transfer of data across international borders.

“Discussing sovereignty is becoming less about meeting regulatory requirements, though those remain important, and more about organisations taking control and maintaining oversight of their data and digital infrastructure,” Harvey explainsed

The company’s approach to the issue of sovereign cloud was also recently spelt out in no uncertain terms by Ashesh Badani, the senior vice president and chief product officer, Red Hat at the aforementioned summit.

“Innovation should not be a trade-off for control. Whether an organisation is meeting jurisdictional mandates or reclaiming its data from proprietary silos, we are providing the capabilities and platforms to build a more self-determined future.

‘’Red Hat is focused on helping the organisations that use these technologies to drive the next decade of AI and cloud innovation on their own terms,” he said.

In response to the move to sovereign cloud and its stance on self-determination being paramount, Red Hat has expanded its sovereign and private cloud offerings, helping global organisations reclaim authority over their technology stacks and data.

New updates include simplified audit compliance, production-ready cloud landing zones, rapid delivery for sovereign AI, and in-region content delivery. The expansion also builds on Red Hat Sovereign Cloud, the brand’s open source foundation for hosting critical sovereign workloads.

“Simplifying sovereignty is an exercise in reducing costs, reinforcing compliance and expediting access to AI and cloud services that are essential for businesses to grow. South Africa is on track to define its rules surrounding sovereignty, and so the path forward for businesses is to transform how they handle data and applications from the start,” Harvey added.

From disruption to certainty: Why business priorities are shifting

One of the many interesting themes to emerge from the discussion was how business priorities appear to be shifting.

Whereas in the previous decade, businesses seemed to be actively seeking uncertainty in the form of disruption and the next wave of innovation, enterprises today seem far more focused on securing some measure of certainty and predictability. That does not mean innovation has lost its appeal. Rather, it suggests that in a fast-changing environment, businesses are increasingly looking for technologies that can help them move forward without adding unnecessary risk or instability.

That idea aligns with Red Hat’s view of the current market. Bruce Busansky, the specialist in Application Platforms at Red Hat South Africa, noted that each major technology wave, from Linux to cloud to containers and now AI, has been defined not only by new possibilities but by the need for a stable foundation underneath them.

In that context, the goal is no longer disruption for its own sake, he stressed. It is to help organisations adopt what is new in a way that is safer, more controlled and more practical, whether that means standardised platforms, support across environments, or hardened systems that reduce the pressure of constant change.

Also, in speaking with Red Hat’s team, open source seems to be the big winner here, with its foundations built on adaptability, freedom and flexibility, all of which are still essential in an age of age uncertainty, and possibly even more so. What really stood out to me in my conversations with the team is people inherently dislike having anything being forced on them, and that includes AI. It seems we are wired to desire to exercise our ability to choose - an ability that AI doesn't (yet) possess, and perhaps that is why open source and self-determination holds such enduring and endearing appeal.

Importantly, certainty does not mean rigidity. Harvey stressed that what enterprises ultimately want is the ability to make decisions that both preserve flexibility and choice, which is yet another balance to be achieved, in an uncertain world. That is where open source remains central: first as a force for freedom in the Linux age, then in hybrid cloud world, and now in an indisputably AI era.

If I am reading the headwinds correctly, then if the last decade was about embracing disruption, this one may be more about building the kind of resilience that allows businesses to innovate with greater confidence, while still retaining their options.

My hunch though is that the biggest balancing act that we collectively face, as individuals, businesses and enterprises, is really about using AI to its full benefit while retaining what makes us uniquely human, and bringing the best of both worlds together.

Enjoyed this? Don’t stop here.

I explore the human side of AI, technology, and the future in a way that’s actually useful.

For a deeper, more personal look at what doesn’t make it into the article:

👉 Join me on Substack (for deep dives)

👉 Join my WhatsApp Group (for real conversations)