As Nigeria shifts towards becoming a more digital economy—from mobile banking and fintech to online commerce—cyber criminals are becoming more active as well as sophisticated in their attacks. They are starting to onboard artificial intelligence (AI) tools, which makes scams easier to launch. For example, using Generative AI (Large Language Models) can help cyber criminals compile more convincing phishing messages on demand that include text, images or video.

“Where old phishing emails were full of typos, now, Large Language Models (LLMs) enable attackers to craft personalised, convincing messages with good grammar, flow, and structure,” says Benjamin Okolie, technology expert and consultant in Africa at Kaspersky. “Fraudsters can even mimic the tone of a known colleague or executive by scraping their social media posts. AI also automates visuals: attackers can seamlessly create striking visuals or even build complete landing pages, making fake banking sites or invoices hard to distinguish from real ones.”

AI is not limited to text. Deepfakes – AI-generated voice and video – add a new dimension to scams. With seconds of audio, criminals can clone anyone’s voice to leave bogus voicemail or WhatsApp requests pretending to be family or a boss, asking for urgent money transfers. Video deepfakes are becoming similarly accessible: fraudsters can swap a face into an existing video and add a realistic voice. In practice, an attacker could conduct a live video call posing as a trusted associate or even a romantic partner.

Okolie warns that deepfake attacks will grow in sophistication. Companies and individuals should prepare.

“We anticipate an increase in targeted attacks using deepfakes, especially against influential people and high-net-worth individuals or organisations,” he adds.

Once confined to science fiction, these threats are now part of the global cybersecurity reality. Kaspersky data shows 893 million phishing attempts blocked in 2024 (up 26% year-on-year), and AI makes them more convincing. Kaspersky experts predict that scammers will amplify their tactics using neural networks, using AI to generate ever more believable fraudulent images and videos.

The human side of security is also under pressure. A recent Kaspersky study found that 51% of employees in the Middle East, Turkiye and Africa (META) region claim they can spot a deepfake, but in a blind test, only 25% succeeded. In other words, many individuals think they are safe, but AI scams can fool even savvy users.

To protect Nigeria’s digital economy, businesses need both next-generation tech and employee awareness. On the technical front, Kaspersky’s latest suite – known as Kaspersky Next – is explicitly designed for the AI era. It is an AI-powered product line that combines advanced endpoint protection (EPP/EDR) and extended detection and response (XDR) capabilities with protection for corporate communications, cloud assets and awareness training for IT specialists.

In practice, Kaspersky Next uses real-time machine learning and integrated detection/response tools to prevent sophisticated attacks across devices, servers, email, and cloud. This top-tier protection helps organisations spot and block AI-driven phishing and intrusions before they succeed.

Equally critical is the human firewall. Well-trained staff can stop scams that slip past tech. Kaspersky’s Automated Security Awareness Platform (KASAP) is an online training system that builds employees’ cybersecurity skills level by level. KASAP offers 350+ interactive lessons and simulations covering every topic from phishing and social engineering to deepfakes. Engaging staff with tests and simulated attacks reinforces the cautious habits businesses need. In Nigeria’s context, regular employee training via tools like KASAP, alongside strong security software, can close the gap exposed by AI fraud.

“By combining powerful, AI-aware security solutions with ongoing staff education, companies in Nigeria can harden themselves against this new wave of scams and benefit from wider use of IT without high cybersecurity risks,” concludes Okolie.