EP 42 | “Sell Intelligence, Not Raw Data” - Mark Nasila on Data Sovereignty

Loading player...
AI must create real value, not just proofs-of-concept. Dr. Mark Nasila explains how FNB’s risk function uses data and AI to stay ahead of evolving regulation and fraud -preventing some R2 billion in losses annually - and why people now make decisions after AI aggregates the evidence (including small language models drafting forensic reports for roughly 220,000 investigations a year). He details how to avoid hallucinations and over-reliance on tools with a clear human-in-the-loop sign-off.

From “African AI” as a value-first agenda (health, energy, local constraints) to the industry’s fixation on use cases over end-to-end transformation, Nasila argues the metric that matters is experience and outcomes, not shiny tech. He calls for national focus areas, public–private partnerships, and data sovereignty that controls the AI value chain -“sell intelligence, not raw data” - backed by urgent, coherent AI strategy, not just policy.

Chapters / timestamps

00:00 – Intro: role of Chief Data & Analytics Officer (risk, regulation, proactiveness)
01:34 – Regulations as trust & value; automating compliance at scale
03:10 – AI in risk: ~R2 billion fraud prevented; 220k investigations; SLM forensic reports
05:24 – Hallucinations & over-reliance: limits of models; human-in-the-loop sign-off
08:34 – Defining “African AI”: value-first, local problems, infra, skills, ethics
10:53 – Find the right problems: R&D and prioritisation inside firms & government
12:23 – Why most are stuck at POCs; transform processes, not demos
16:16 – “AI washing”: inflated POC claims vs measured, experience-level value
19:23 – National AI policy: prioritise sectors; augment jobs; socio-economic lens
22:42 – People still matter: limits of chatbots/coding; bring humans back where needed
26:04 – Government data, strategy, and global benchmarks (US/EU/China focus & investment)
28:49 – Efficiency examples and why strategy must lead policy
31:07 – AI as geopolitics/industry driver; sovereign AI and owning the value chain
38:06 – Data centres vs sovereignty; “sell intelligence,” protect the pipeline
41:13 – Where to find the books; closing remarks & next steps

#TheAnglePodcast #MarkNasila #FNB #AI #FraudPrevention #RiskAnalytics #DataSovereignty #HumanInTheLoop #AfricanAI #AIAtScale

Produced and Published by Submedia.co.za
14 Jan English South Africa Technology · Business

Other recent episodes

EP43 | Build Your Brand, Then Scale: Marko Stavrou’s Founder Guide

Entrepreneur Marko Stavrou (“the Gen Z guy”) talks about starting Gen Link after a Standard Bank brief to connect young people to the bank’s ecosystem, and how the company has since worked with 50+ clients. He explains why many organisations overlook youth (low current spending power, short-term optimisation), and argues…
21 Jan 21 min

EP 41 | Every Industry Will Be Disrupted - Celiwe Ross on What Leaders Must Do Now

Leadership in a disrupted world. Celiwe Ross - South Africa’s first black woman to qualify as a mining engineer reflects on leading through technological change: self-knowledge over fear, community over ego, and why yesterday’s leadership model won’t work when we’re partnering with AI. She shares early underground lessons on teamwork…
7 Jan 29 min

EP 40 | Pay-on-Proof AI: How Isazi.ai Turned Clients into Evangelists

Practical AI beats hype. In this Singularity Summit episode of The Angle, Ashley Anthony - co-founder & CEO, Isazi.ai explains how a bootstrapped AI company grew to 110 people by solving specific customer problems first—then choosing the tech. He unpacks Isazi’s first-principles approach - translate business issues into mathematical problems…
31 Dec 2025 33 min

EP 39 | Keeping African Agency in AI: Vukosi Marivate’s Call to Build

Keep African agency in AI - that’s the through-line of Prof. Vukosi Marivate’s talk. Don’t outsource the future, build it locally. He points to a decade of grassroots work (Data Science Africa, Deep Learning Indaba, Masakhane) and argues for a full ecosystem: fundamental researchers, model builders, product teams, and real…
24 Dec 2025 26 min

EP 38 | Speed, Open Source, Ecosystems: The Faster Path to AI Value

How do you build AI that serves profit and purpose—without causing unintended harm? In this episode of The Angle, Atenkosi Ngubevana - Group AI & Automation lead at Vodacom unpacks a values-first approach to AI: design for inclusive growth, embed ethical data stewardship, and measure impact beyond demos. We get…
17 Dec 2025 38 min