From Generative to Agentic AI: Why the Next Phase May Be Africa's Biggest Tech Opportunity

The generative AI wave moved faster than almost any technology in modern economic history — and yet, for many businesses and governments, it arrived before they fully understood what they were adopting. Now, with agentic AI emerging as the next phase, the question of who benefits and who falls behind becomes more urgent, particularly for economies that have historically arrived late to foundational technology transitions.

Johnson Idesoh, Group Chief Information and Technology Officer at Absa, argues that Africa doesn't have to repeat that pattern — and that the conditions that enabled the continent's mobile money revolution may be the same ones that allow it to leapfrog into agentic AI in meaningful ways.

Generative vs. Agentic: The Core Distinction

Where generative AI responds to a prompt and produces an output, agentic AI pursues an objective. Rather than completing one task at a time, an agentic system determines the steps needed to reach a broader goal — identifying which tools to use, which sub-tasks to prioritize, and how to adapt when the initial approach doesn't work — with minimal human intervention between steps.

Key Takeaways

• The leapfrog precedent: Africa's mobile money ecosystem didn't follow traditional banking pathways — it grew from constraints, not despite them. Idesoh argues the same dynamic could apply to agentic AI adoption across financial inclusion, fraud detection, and small business operations.
• The risk of downstream adoption: Without deliberate investment, there's a real danger that African organizations become consumers of AI systems designed elsewhere, accumulating less economic value from the transition than the regions and companies building foundational models.
• What's needed: Early investment in people and governance, locally grounded applications that solve African problems — and stronger coordination between banks, regulators, telecoms, and policymakers as autonomous systems integrate into everyday economic activity.

For business leaders in any market, the framing here translates: AI adoption strategies designed around specific operational problems — not broad platform experiments — are more likely to deliver durable value as agentic systems mature.

🔗 Read the full article on CIO Africa