While the global race to harness artificial intelligence (AI) accelerates, a new Cisco report shows that only 18% of South African organisations are truly AI-ready — but those that are, the so-called “pacesetters”, are already reaping the rewards. According to Cisco’s 2025 AI Readiness Index, these companies are four times more likely to move AI pilots into production and 50% more likely to report measurable business value.
The findings, released ahead of Cisco Connect, which took place in Johannesburg on 23 October, paint a revealing picture of South Africa’s readiness to compete in the era of intelligent automation. While 93% of local organisations plan to deploy AI agents — with nearly half expecting them to work alongside employees within a year — most lack the secure infrastructure and data discipline needed to sustain AI at scale.
“AI is accelerating change, no matter the industry,” says Smangele Nkosi, General Manager of Cisco South Africa. “AI value doesn’t come from experimentation alone; it comes from execution.”
Turning pilots into profits
Cisco’s study, based on feedback from over 8,000 global AI leaders, identifies a consistent elite group of “pacesetters” — roughly one in five South African organisations — who outperform peers across every metric of AI maturity. These firms share one crucial trait: they treat AI as a strategic capability rather than a side project.
Nearly all of them (99%) have a defined AI roadmap compared to just 59% of South African companies, and 91% have a formal change management plan to support adoption. Their budgets reflect that intent — 79% rank AI as their top investment priority versus only 23% locally.
What truly sets the leaders apart, however, is execution discipline. More than 60% of pacesetters have a repeatable process for scaling AI use cases, and three-quarters have already finalised them. Almost all track ROI and productivity metrics, and 71% expect to generate entirely new revenue streams from AI initiatives.
Infrastructure debt: the hidden drag
The report also introduces a critical new concept: AI Infrastructure Debt. Similar to technical debt, it refers to the silent accumulation of outdated systems, deferred upgrades and underfunded architecture that can erode the value of AI investments over time.
Cisco’s research shows that 40% of organisations expect workloads to rise by at least 30% in the next three years, yet only 23% have sufficient GPU capacity to manage those demands. Another 64% struggle to centralise data, while two out of five lack the ability to detect or prevent AI-specific threats.
For many, the result is an AI ambition that far exceeds operational readiness. Nkosi warns that without secure, scalable systems, the promise of AI could quickly turn into a burden of complexity and risk.
Agents of disruption
As generative and agentic AI systems evolve, the pressure on infrastructure, governance and trust will only increase. While enthusiasm for AI agents — autonomous systems that can think, act and learn — remains high, the report suggests South Africa’s readiness to support them is still maturing.
Just 18% of local organisations describe their networks as flexible enough to scale for AI complexity. By contrast, 98% of global pacesetters are already designing networks specifically for AI growth, scale and security.
For these leaders, readiness is not about hype but about architecture. They are investing early, aligning strategy with data pipelines and security frameworks to ensure AI deployments generate sustainable, measurable value.
Value follows readiness
Cisco’s AI Readiness Index ultimately argues that the race to AI value is no longer about experimentation but about foundation. The most AI-ready companies are proving that discipline — not just curiosity — drives long-term advantage.
For South African businesses, the message is clear: AI readiness is the new competitive currency. Those that build for resilience today will lead the market tomorrow.