The next phase of artificial intelligence (AI) will be defined by autonomous agents, human-AI collaboration and Africa’s ability to build technology ecosystems suited to its own realities.
This came to light at ITWeb’s AI Summit 2026, where Johnson Idesoh, group chief information and technology officer at Absa Group, delivered a closing keynote titled: “What comes after GenAI? A look at emerging technologies like agentic AI, autonomous AI agents and their potential for South Africa”.
He pointed out that the next phase of AI will move beyond GenAI into more autonomous, agent-driven systems, but stressed that human oversight will remain central.
While AI adoption in SA is accelerating, with some organisations moving from early experimentation into active implementation, Idesoh said AI is an emerging technology, but one with enormous transformative potential for SA and the African continent.
“We are still at the stage where this is such a young technology that, actually, in the long-term, AI integration will become very visible. I'm an optimist. I actually think AI is a great equaliser for South Africa and Africa,” noted Idesoh.
He explained that despite the evident socio-economic challenges, he is broadly optimistic about SA and Africa’s prospects, pointing to the continent's young, mobile-first population and rapid technology adoption as key advantages.
He cautioned that most large language models are still trained on non-African data, reinforcing the need for localised innovation, stronger trust in institutions, and responsible deployment of AI to ensure inclusive economic participation.
Beyond GenAI
Idesoh framed GenAI as only the beginning of a broader technological evolution, arguing that the industry is already moving towards AI agents capable of acting more autonomously.
“Today I'd say it [GenAI] generates new content for users, and they can ask it any question. Generated content can be pretty impressive in some cases. The first version of my slide deck, by the way, was generated by AI, of course with a lot of human work and editing to produce the actual thing.”
He noted the next stage of GenAI development is increasingly focused on agents that can interact independently with systems and even with each other.
“We're very clear that anything that agents do, humans are in control. Personally, I don't think that's going to change. You need humans in the future and then of course there are commentators talking about autonomous AI agents that start engaging with each other.”
Idesoh said Africa’s demographic profile, mobile-first economy and growing digital infrastructure position the continent to benefit significantly from the next wave of GenAI innovation.
“We are the youngest continent on the planet. Young people's adoption of this technology is very, very high and this is part of my optimism. In Africa, the rate of adoption of this technology is not lagging the world, far from it.”
He compared AI’s potential in Africa to the rise of mobile money, arguing that the continent has historically developed technology solutions tailored to local realities rather than simply copying international models.
“Mobile money didn't emerge on our continent because we were copying what other countries were doing. Africa’s strong adoption of mobile technology creates fertile ground for AI-driven services. AI will thrive in a mobile-first world. We use cloud-based architectures and API-driven architectures. You will all have seen increasingly the big hyperscalers investing in South Africa.”
AI and financial services
As technology leader at Absa Group, Idesoh highlighted several practical AI applications already being used inside the banking environment, particularly around fraud detection and cyber security.
“Of course, we use AI here because simply the rate at which fraud is happening means the chances of a human being able to summarise these patterns is next to zero. Another example we've got is we have an agent in our cyber security team. You see something suspicious, send it in, it analyses it quickly and it comes back with results.”
He explained that AI systems are helping the bank process suspicious activity at a scale impossible for human teams alone.
“I could not employ enough people to look at all the phishing e-mails and suspicious e-mails sent every day. A human can look at these things probably for hours and hours. The agent does it in two minutes.”
Despite enthusiasm around AI, Idesoh stressed that organisations must deploy the technology responsibly and ensure it expands access rather than deepening inequality.
“This is powerful tech. We need to make sure we use it to break down barriers. Inclusion has got to be our mission. This thing about organisational trust is really important. We must keep building trust in our institutions because these are all the things that attract investment. The future is where humans and these tools are working together,” he concluded.
Share
