Knowledge Hub

The world is entering a defining moment for digital infrastructure. Artificial intelligence has moved from experimentation to ubiquity, and with it, a new operational paradigm is taking shape — one where agents rather than applications become the primary engines of digital value creation. This is the dawn of the agentic AI era, and its success depends on one thing above all else: robust, intelligent, and scalable infrastructure.

It has become a common refrain – getting data governance right is key to a successful AI strategy! This conventional wisdom is very true, but it is not a new problem. For as long as I have been involved in IT, both as an analyst and as a CIO, companies have struggled with wrangling the various data sets across the applications running at the organization.

Organizations have invested billions in advanced cybersecurity tools like firewalls, SIEM platforms, and zero-trust architectures. Yet, breaches persist. The root cause is not a technology gap but a human challenge. Mimecast’s The State of Human Risk 2026 report, based on input from 2,500 IT professionals across nine countries, highlights this stark reality: organizations know their vulnerabilities but aren’t acting quickly enough to address them.

South African CIOs are entering the agentic era under conditions that leave little room for experimentation without purpose. AI spending in South Africa is rising quickly, but organizational maturity remains early: IDC projects local AI-related spending to exceed $7 billion by 2030, yet more than half of South African organizations still sit in an ad hoc, experimental stage of AI maturity today. That tension matters. It means the opportunity is real, but so is the risk of fragmented investment, inflated expectations, and stalled value realization.

To understand today’s challenge, it helps to look at how technology investment has evolved.

Spreadsheet Economics: Static, manual, and predictable. Annual planning and periodic reconciliation provided a sense of control.
Cloud Economics: Consumption-based and dynamic. FinOps emerged to manage distributed and fluctuating costs.
AI Economics: Non-linear, decentralised, and rapidly scaling. Every team builds and consumes independently.

The era of centralized AI is ending. What began as a race to adopt single-cloud AI services is giving way to a more complex—and far more powerful—reality. As AI becomes embedded in every business process and decision, enterprises are confronting a simple truth: intelligence cannot live in one place. Data doesn’t. Users don’t. Regulation doesn’t. Competitive advantage now depends on how effectively AI is distributed across the digital landscape.

AI is entering a new phase. Organizations are moving beyond experimentation and looking at how intelligent systems can support day-to-day operations, improve decision-making, and help employees work more effectively. The rise of agentic systems AI tools capable of acting with greater autonomy and context awareness – is accelerating that shift and redefining the CIO agenda.

The first wave of AI gave organisations new tools — ways to generate content faster, analyse data at scale, and automate the repetitive. Useful, certainly. But the underlying work model remained largely intact. Humans still decided, directed, and executed. AI assisted.