Business leaders must shift from a "forgiveness over permission" mindset to an "ethics-by-design" approach to unlock the full potential of artificial intelligence in Africa.
This was the core message from the ITWeb AI Summit 2026, where industry experts warned that retrofitting trust into already-deployed AI models is often impossible and prohibitively expensive.
Tahir Latif, chief trust officer for EMEA at Securiti AI (a Veeam company), told delegates that while big tech firms in the US often bake regulatory fines into their profit and loss statements, most organisations cannot afford the $5 billion penalties associated with global breaches. Latif argued that establishing ethical guardrails at the start of innovation is a business necessity, not an academic exercise.
He highlighted that deploying AI quickly may offer short-term gains, but the downstream risks – including data corruption and privacy violations – can be catastrophic.
"Wait until you see the downstream risks that are invisible until they materialise," Latif cautioned. "The real cost lies in trying to retrofit trust, security and ethics into models already deployed."
Johan Steyn, founder of AIforBusiness.net, raised concerns about the enforcement of AI regulation in SA. He noted that while laws like the Protection of Personal Information Act carry 10-year prison sentences, literacy and consistent enforcement remain significant challenges.
A major technical focus of the discussion was data sovereignty – the principle that data is subject to the laws of the country where it is collected. Latif clarified that sovereignty is not just about where data sits, but who can access it. In highly regulated sectors like banking, data often cannot leave SA, yet it is frequently processed in Southeast Asia or backed up in North America without proper oversight.
The emergence of agentic AI – autonomous systems that act to achieve business goals – presents a new frontier of risk. These agents often exceed their remit, touching or dropping sensitive data sets they were never intended to access.
Latif suggested that as AI becomes a commodity, trust will become the primary differentiator for African businesses. Companies that can prove an audit trail of ethical data handling – rather than merely checking a training box – are likely to win in the long-term digital economy.
Share
