
Samuel Sikazwe, Garden Lusaka, 18 November 2026–In the past, poor grammar and awkward phrasing gave away fraudulent payment requests before money left the account. A familiar voice, a shared language and a known relationship were proof of identity until attackers began cloning voices and copying writing styles in local languages. A call-back verification check made through a separate channel can now confirm whether an urgent payment request is genuine, and it is one of four controls outlined by Ms Patricia Nyati, Head of Legal and Regulatory at Liquid Intelligent Technologies Zambia.
A Voice That Used to Settle the Question
Payment requests and bank detail changes need call-back verification. A person who receives an instruction by phone or email confirms it by contacting the sender through a different channel. Three companion controls complete the set of four. Transfers need approval from two people, multi-factor authentication asks for a second proof of identity at login, and staff work in a culture that welcomes questions about urgent requests. Ms Nyati explained the change that makes them necessary.
“AI has removed the tell-tale signs we used to teach people to look out for, such as poor grammar and awkward phrasing,” she said. “Attackers can now generate polished, personalised messages at scale, mimic the writing style of a real executive, and even clone voices or create convincing videos for payment-approval fraud.”
Most Zambian enterprises are not yet fully prepared. Awareness training tends to be annual and generic, while many payment processes still rely on a single email or phone instruction. A cloned voice that sounds like a trusted senior colleague is hard to counter through awareness alone, so training must sit beside firm process controls.
Why Zambian Businesses Draw Attention
Ms Nyati named the main threats and why they have multiplied. “Phishing, business email compromise, ransomware, identity theft and attacks on suppliers are particular concerns,” she said, “and that is because Zambian and African businesses have digitalised quickly through mobile money, online banking, cloud tools and remote work, which has widened the attack surface faster than the pace at which security capability has grown.”
African enterprises draw attackers for three reasons, she said. Digital payments are growing quickly, customer datasets hold value, and cybersecurity maturity varies from one institution to the next. Financial services, telecommunications, healthcare, government and other organisations that handle large volumes of customer or payment data face significant exposure. Smaller enterprises carry their own risk, because security spending has trailed the pace at which they adopted digital tools.
Local knowledge sharpens each attack. Criminals tailor mobile-money fraud to local habits, impersonate executives and suppliers, send fraudulent changes to payment instructions and run scams on messaging platforms. Zambian languages, regional business practices and trusted relationships make global techniques more convincing.
The Quiet Middle of an Attack
A modern attack rarely announces itself. It begins with a convincing phishing message, a stolen password or a remote access service left exposed. Once inside, intruders explore the network, widen their access, locate sensitive data and often steal it. Attackers disable or delete backups before they deploy ransomware or threaten to publish what they took, and they usually strike outside normal business hours. By the time anyone notices, the breach has often been underway for days or weeks.
In Ms Nyati’s experience, most organisations miss the intrusion in the middle of this sequence. “Nobody is watching for unusual logins, new administrator accounts, large data transfers or security tools being switched off,” she warned.
She traced that blind spot to practical gaps. Multi-factor authentication is often missing on email and remote access, software patching is delayed, logs go uncollected or unread, and without round-the-clock monitoring early warning signs go unseen.
Where Cloud Moves the Risk
Ms Nyati addressed a common worry about cloud services. “Cloud itself is not less secure, but it changes where the risk sits,” she countered.
The problems she sees most often include storage left open to the internet, accounts that hold more privileges than their owners need, weak identity controls, poor visibility across several cloud services and a misunderstanding of shared responsibility. The provider secures the platform, and the customer remains responsible for data, access and configuration. Security strategy therefore centres on protecting identities, data and workloads, using identity and access management, multi-factor authentication, continuous monitoring of configurations, encryption and centralised logging. Where data is hosted also matters, since residency requirements and obligations under Zambian data protection legislation apply.
Habits That Put Preparation First
Ms Nyati sorts organisations into two groups by how they prepare. “A reactive organisation asks what to do after an attack,” she said. “They wait for something to break, then call for help, while a proactive organisation has already decided how to detect, contain, recover and learn from previous incidents.”
Preparation starts with mapping the assets that matter most. Regular risk assessments, prompt patching, threat monitoring, tested backups, rehearsed incident responses and trained staff follow, along with scrutiny of supplier security and a continuous check that existing controls work. A plan written in advance settles who decides, how evidence is preserved, when specialists are called and who must be told under the law. Ms Nyati listed the most damaging mistakes as acting without a plan, delaying escalation, failing to isolate affected systems, deleting evidence, sharing unverified facts and restoring systems too early. Paying a ransom guarantees neither restored systems nor recovered data.
Ms Nyati described how Liquid Intelligent Technologies Zambia structures a typical engagement. It begins with an assessment of the customer’s environment, risks and existing controls. Liquid then connects those systems to a central monitoring platform with alerts tuned to normal activity. Specialists monitor for threats and support investigation, containment and recovery, and regular reports and recommendations keep security improving over time. The aim is a long-term partnership in which customers gain capabilities that would be difficult and costly to build in-house.