South African organisations need to shift from AI experimentation to measurable business outcomes, trusted governance and scalable execution, according to Lee Naik, CEO of TransUnion Africa.
Speaking at the ITWeb AI Summit in Johannesburg on Wednesday, Naik said too many enterprises remain caught up in the excitement around generative AI while failing to address the country's economic challenges, including unemployment, cost pressures and sluggish growth.
Debates over which large language model is superior are irrelevant if organisations cannot use the technology to solve real problems, he told delegates.
“AI is steadily fading into the background and becoming embedded in everyday systems, processes and workflows," Naik said. "Businesses need to stop treating AI as a standalone trend and start seeing it as core infrastructure.”
He warned that many companies invest in pilots that never progress to production. “These projects often absorb budget and management attention without producing any measurable return.” Naik referred to these stalled initiatives as “zombie pilots” and said organisations should be willing to shut them down if they cannot demonstrate commercial value.
Naik said technology teams often undermine themselves by presenting AI projects in technical language focused on models and architecture, while executives look for revenue growth, margin improvement or cost savings. “The obligation is on technologists to learn the language of business,” he said.
To help delegates assess AI readiness, Naik proposed a three-part framework built around value, factory and trust.
Value means clear, measurable business impact. If business leaders do not perceive value from an AI initiative, the project has failed regardless of its technical sophistication.
Factory refers to the ability to industrialise AI delivery so successful use cases can be repeated consistently. AI must become an operational capability, not a side project.
Trust requires confidence that models are fair, explainable and properly governed – especially in financial services, where automated decisions affect access to credit and fraud prevention.
Naik also highlighted financial inclusion as a significant AI opportunity. “Millions of Africans remain excluded from formal credit markets because they lack conventional financial histories,” he said. “By using alternative data sources and advanced analytics, lenders can extend responsible credit to consumers who would otherwise be overlooked.”
Naik urged organisations to review existing AI initiatives, identify projects that solve genuine pain points and scale only those that produce measurable outcomes. The period of AI novelty is ending, he said. The organisations that will benefit most are those that combine commercial discipline, governance and execution.
Share
