Episode 188 - From Data-Rich to Insight-Driven: How AI Is Reshaping Retail in Kenya & Africa

Episode 190 - WhatsApp Usernames, Privacy, Fraud & Digital Identity: An Interview with the BBC's Nkechi Onyinyechi Ogbonna on the Focus on Africa Podcast

WhatsApp is preparing to introduce usernames that will allow people to communicate without publicly sharing their phone numbers. The feature could deliver a meaningful privacy improvement, particularly in group conversations, business interactions and chats with people who have only recently met. However, it has also raised questions about impersonation, phishing, financial fraud, mobile-money security and whether accounts involved in criminal activity will remain traceable. On the 15th July 2026, I joined Nkechi Onyinyechi Ogbonna on BBC World Service’s Focus on Africa to discuss WhatsApp’s forthcoming username feature. The programme also featured Somalia’s State Minister of Communications and Technology, Hon. Ahmed Osman Diiriye, who explained why his government wants Meta to provide stronger assurances around fraud, mobile-money security and digital traceability. The discussion examines why Somalia wants Meta to demonstrate “proof of traceability,” the concerns India has raised and the protections WhatsApp says it is building into the feature. It also explains: * The difference between reserving and activating a username * How usernames differ from display names * Why users will still need telephone numbers * How optional username keys will work * Why WhatsApp will not offer a public username directory * How usernames could protect people in group chats * What the feature means for creators and businesses * The risks of lookalike usernames and impersonation * Why mobile-money markets require additional safeguards * How digital literacy can help users identify scams Chapters * 00:00 — Introduction to WhatsApp usernames * 00:55 — Somalia’s security and accountability concerns * 01:53 — Consultations between Somalia and Meta 02:16 — Is there evidence that usernames will increase * fraud? * 03:22 — “We want proof of traceability” * 04:27 — Meta explains its proposed safeguards * 05:19 — Why WhatsApp is separating identity from phone numbers * 06:26 — How usernames will change one-to-one and group communication * 07:11 — Impersonation, financial fraud and mobile-money risks * 09:02 — Digital literacy and practical user protection * 10:50 — Conclusion

Episode 189 - Africa’s Scam Economy Has Industrialized: What Telecom Operators Must Do Next

Africa’s scam economy is becoming more organised, automated and sophisticated. What was once the occasional fraudulent SMS has evolved into smishing, vishing, identity theft, business impersonation, SIM-swap-enabled account takeovers, AI-generated identity documents and deepfake voice calls. I moderated a discussion recorded at the TARS 2026 Telecom Africa Revenue Assurance & Fraud Management Summit in Nairobi, Kenya, on the 13th May 2026 titled: Protecting the Customer: Scam Typologies, Operator Responsibilities & Fraud Awareness in Africa The panel brought together: * Ann Khambo - Revenue Assurance and Fraud Management Manager, Airtel Money, Airtel Kenya * Ogochukwu (Ogo) Onwuzurike - Country Manager, Nigeria, Truecaller * Mieraf T. Birhane - Executive Head of Fraud Management, Safaricom Telecommunications Ethiopia The discussion examines how customer-facing fraud is changing across African telecom and mobile-money markets, why social engineering remains such an effective attack vector, and why operators can no longer regard themselves as neutral pipes when their networks, brands and identity systems are being exploited. Mieraf shares insights from Ethiopia’s evolving fraud landscape, including smishing, account takeover, subscription fraud and commission-related abuse. She also describes a striking case in which fraudsters used makeup and physical impersonation to attempt fraudulent SIM swaps—an incident detected because the agent involved had received practical fraud training. Ann explores identity theft, including situations where innocent customers may not know their photographs or personal information have been used to register SIM cards or commit fraud. She also explains why GSM and mobile-money fraud must be analyzed together using KYC, voice, SMS, location, SIM, device and transaction data. Ogo introduces the “machine era of spam and fraud.” According to the figures shared during the panel, Truecaller intercepted approximately 68 billion spam and fraud calls in 2025, compared with approximately 37.8 billion in 2021. She explains how AI and automation are helping fraudsters scale while excessive calls and messages from legitimate organisations are also eroding trust in phone-based communication. Key Themes: * Smishing, vishing and social engineering * Identity theft and fraudulent SIM registration * SIM-swap-enabled account takeover * Business impersonation and investment scams * Real-time detection, analytics and machine learning * Internal fraud and insider risk * Anti-fraud by design * Operator responsibility and customer protection * Cross-industry collaboration Time Stamps 00:00 Intro music 00:12 Africa’s scam operations have industrialised 04:09 What keeps fraud leaders awake at night? 04:44 Ethiopia’s evolving fraud landscape 08:43 Identity theft and the innocent customer 10:54 The human factor in customer-facing fraud 11:24 Why operators must take greater responsibility 13:43 The machine era of spam and fraud 18:05 AI call scanning and verified business communication 19:40 Smishing and sophisticated account takeover in Ethiopia 23:21 Data-driven controls for mobile-money fraud 24:05 Connecting GSM and mobile-money fraud 27:52 Business impersonation, investment scams and credibility engineering 30:49 Outro music The central takeaway is that customer-facing fraud is no longer only a fraud-management or revenue-assurance issue. It is a customer-experience, brand-trust, financial-inclusion and digital-economy issue. Protecting customers requires shared responsibility across telecom operators, mobile-money providers, banks, fintechs, platforms, regulators, agents, employees and customers.

Episode 188 - From Data-Rich to Insight-Driven: How AI Is Reshaping Retail in Kenya & Africa

Kenyan and African retailers generate enormous volumes of information through point-of-sale systems, inventory platforms, loyalty programmes, e-commerce, mobile applications, social commerce, delivery platforms and digital payments. However, much of that data remains fragmented, underused and disconnected from everyday business decisions. In this special episode of the Pure Digital Passion Podcast, I moderate a discussion from the RETRAK Retail Summit 2026 titled: Smarter Retail: Turning Data, AI and Insights into Competitive Advantage The session took place on Thursday, 14 May 2026, at the Sarit Expo Centre in Nairobi and brought together four experts representing retail entrepreneurship, digital payments, financial-services platforms and retail technology, as follows: * Judy Waruiru - Regional Managing Director, Network International * Sonal Haria - Co-Founder and CEO, Canvas Cosmetics, Co-Founder, CB Consulting & Media Group * Eric Muriuki - Group Director, Digital Business, CEO, LOOP DFS, NCBA Group * Siddesh Narkar - Head of Product, Compulynx Topics Covered * Why many retailers are data-rich but insight-poor * The importance of connecting fragmented customer data * Building a unified view across stores, websites, apps and payments * Why data readiness must come before AI readiness * Using data to improve product development, pricing and assortment * How Canvas Cosmetics used customer insights to guide product development * Using payments data to understand churn, market movements and fraud * AI-supported replenishment, stock management and pricing * How AI can improve sales productivity and outreach * Balancing personalisation with customer privacy and trust * Privacy-by-design approaches to retail data * How AI agents may soon shop and pay on behalf of consumers * Practical AI actions retailers can take during the next 12 months Key Message * Retailers do not need to begin with an expensive, enterprise-wide AI programme. * They should begin by digitizing operations, organising existing information, selecting one or two commercially important problems, measuring the results and building from the small wins. * AI can analyze more information and provide options faster, but human judgement remains essential. Chapters 00:00 Introduction to the RETRAK panel00:44 The state of retail in Kenya and Africa07:07 Sonal Haria on combining data with human judgement09:20 Judy Waruiru on fragmented customer data10:43 Eric Muriuki on data readiness and AI15:50 Siddesh Narkar on clean and usable retail data17:32 How Canvas Cosmetics used data for product development21:12 Payments data, customer journeys, churn and fraud26:39 AI as an intelligent wrapper around the business33:53 Audience questions begin34:36 Preparing and labelling business data for AI38:25 AI-generated cosmetic formulations and human oversight39:44 Understanding wider market and industry trends40:48 Using payments data for retail-market intelligence42:07 Turning financial-service providers into insight partners44:59 How AI improved sales productivity and outreach46:33 AI personalization, privacy and customer trust49:15 Cloud and on-premise AI deployment50:08 Agentic commerce and the machine as the next customer53:27 One action retailers should take in the next 12 months57:42 Closing remarks The Pure Digital Passion Podcast explores the people, organizations, technologies and ideas shaping digital transformation, marketing, media, innovation and business across Kenya and Africa. Subscribe for more conversations with African technology leaders, entrepreneurs, executives, policymakers and innovators. #Retail #RetailTechnology #ArtificialIntelligence #AI #DataAnalytics #DigitalTransformation #Payments #Fintech #CustomerExperience #Kenya #Africa #PureDigitalPassion

Episode 187 - The Future Of Lending In Kenya & Africa: AI, Alternative Data, Digital Identity & Automation

What does a truly future-ready African lender look like? Is it enough to launch a mobile application, automate loan approvals or introduce artificial intelligence into the credit-scoring process? Or does meaningful digital transformation require lenders to rethink the entire customer journey—from acquisition, onboarding and identity verification through credit decisioning, disbursement, repayment and collections? In this panel discussion I moderated a practical and thought-provoking conversation on the future of lending in Kenya and Africa. The discussion was recorded during The Future of Lending: Loan Origination, E-Sign & AI, held on the 29th of May 2026 at Park Inn by Radisson in Westlands, Nairobi. The event was co-hosted by Presta Technologies, Zoho and the Digital Financial Services Association of Kenya. Panelists * Kris Senanu: Executive Chairman, Smith & Berkeley LLC * Kevin Mutiso: CEO, OYE and Chairman, Digital Financial Services Association of Kenya (DFSAK) * Winnie Chira: Founder and CEO, Identify Africa * Victor Kiplagat: CEO and Co-Founder, Spin Mobile LLC * Kenneth Mantu: Group CEO, The Adaptis Group Key Topics Covered 1. Why many African lenders still operate through fragmented platforms 2. The difference between having digital channels and having a genuinely digital lending operation 3. Why reliable data matters more than institutional gut instinct 4. How incomplete information can cause both financial exclusion and over-indebtedness 5. Digital identity, stolen documents, deepfakes and onboarding fraud 6. Risk-based KYC and creating seamless journeys for genuine customers 7. How alternative data can improve decisions for thin-file borrowers 8. Mobile-money transactions and behavioural credit indicators 9. Why correlations in lending data must be interpreted responsibly 10. Embedded finance and the importance of loan purpose 11. Why borrowers value speed, convenience and certainty 12. The local shopkeeper as an overlooked source of informal credit intelligence 13. The role of automation in removing repetitive manual work 14. Why change management is critical to successful technology adoption 15. Practical AI integrations in lending 16. Anomaly detection, model monitoring and human oversight 17. What lenders should prioritize over the next twelve months Chapters 00:00 Unified Lending Platforms, Market Readiness & Credit Infrastructure01:33 Data Integrity & Why Analytics Can Challenge Gut Instinct03:25 Building Sustainable Lending Businesses Through Data05:22 Lending Silos, Manual Workflows & Kenya’s Mortgage Gap07:51 Deepfakes, Stolen IDs & Risk-Based KYC08:48 Customer Acquisition, Fragmented Data & Over-Indebtedness12:26 Digitising Lending Without Overwhelming the Organisation15:01 Speed, Paperwork & the Signs of an Outdated Lender17:47 Don’t Give a Human a Robot’s Job18:08 Alternative Data, Mobile Money & Credit Scoring20:16 Tithing, Loan Stacking, Betting & Affordability Signals22:49 Audience Question-and-Answer Session Begins24:14 Scoring Thin-File Customers Using Feature Phones25:23 Embedded Finance, Loan Purpose, Speed & Convenience28:59 Why the Shopkeeper May Be Africa’s Largest Lender31:20 Can Lifestyle Patterns Predict Borrower Behaviour?31:54 Alternative Data for Collections, Skip Tracing & Product Development34:02 Questions on Scoring Bias, Fraud & Regulatory Complexity35:42 Industry Collaboration Against Fraud40:12 KYC, SIM-Swap Checks, Identity Matching & Document Verification43:04 The One Change Every Lender Should Make43:27 Unified Platforms & the End of Lending Silos44:12 Intelligent Automation Driven by Data44:22 Digitisation & Real-Time Management Visibility44:36 Integrating AI Into Existing Lending Systems45:29 Anomaly Detection, AI Monitoring & Human Oversight47:03 Final Takeaways: Data, Technology & People

Episode 186 - CTRL + ALT + HUMAN: What a Room Full of Business Leaders Taught Me About AI and the Future of Work in Kenya & Beyond

What happens to the humans when the machines can increasingly do the work? Recorded live at the Ikigai Industry Nights event in Nairobi on the 2nd July 2026, this special episode of the Pure Digital Passion Podcast brings you the complete CTRL + ALT + HUMAN panel — as moderated by me (Moses Kemibaro), with Shikoli Makatiani (co-founder & CTO, Akili AI), Victor Ambuyo (Head of Growth, Madavi) and Marvin Oyoo (Management Systems Consultant, Panoramic Synergy) as my panelists. From Shikoli’s 10/90 rule and the loan-intake process that was 90% broken, to Victor’s anatomy of the failed rollout (‘a people problem wearing a technology costume’) and the fluency answer to the jobs question, to Marvin’s reality gap, ‘that’s not a strategy — that’s a subscription’, human–AI synergy and a council of AIs checking each other — plus lamplighters, coexistence, tea-buying drones, ATM forensics and a physics exam passed with an AI tutor. This is practical AI, Kenyan edition: no hype, real examples, honest answers. Recorded before a packed live audience. Take the free Akili Snapshot and find out what AI is doing with your data — in 15 minutes: assured.akili-ai.com Timestamps * 00:00 Welcome · why this conversation, why now — ChatGPT’s 100M users in two months vs Spotify’s eight years · meet the panel * 03:33 Opening round: what the AI hype gets most wrong — and the one shift every organisation must make in 12–18 months * 05:25 Victor: the hype is about tools — the gap is the organisations and people meant to use them * 06:27 Marvin: AI replaces repeatable tasks, not people — the leadership-readiness question * 08:58 Shikoli: ‘Only 10% of the work is AI’ · the loan-intake story · automating decisions, not just processes * 12:52 What’s production-ready in Kenya today vs what’s still a demo * 17:05 Victor’s anatomy of a failed rollout: the event, the subscriptions, the 20% — ‘a people problem wearing a technology costume’ * 21:15 Marvin: the reality gap, live — ‘AI is fun until the real work starts’; the SLM/LLM show of hands; using 5% of what you pay for * 23:32 ‘That’s not a strategy — that’s a subscription’ · AI is everyone’s responsibility, not an IT project · the sensitive-data warning * 25:38 The jobs question I: Victor on judgement, pattern machines, Kenya’s trust economy — and fluency: ‘it’s someone more fluent with AI who takes your job’ * 31:06 The jobs question II: Shikoli — lamplighters, the work that disappears · the hospital insurance example · the red line on human life (‘…I’ll eat the leaf’) * 35:16 From collaboration to coexistence — the Ethan Mollick frame * 35:58 Marvin: human–AI synergy — the human checker, hallucinations and data poisoning · the SOC alert-fatigue example * 40:47 Transparency, consent and bias · why a 50-page policy defeats a model · the ‘AI council’ — models judging models, a human above them * 44:06 Audience Q&A * 45:39 Shikoli: AI’s sleeping superpower — vision: tea drones buying crop months before auction · 300 ATM videos in minutes * 49:46 First principles: AI as thought partner, feedback and tutor — the physics exam · organisations = workflows = tasks * 53:30 Final round + the parting shot: the solopreneur billion-dollar company · close