Nigeria’s cybersecurity debate is shifting beyond hackers and stolen passwords as deepfakes, synthetic identities and increasingly autonomous AI systems expose new weaknesses in the continent’s digital economy.

Artificial intelligence is giving African businesses new ways to serve customers, process payments and automate operations. But the same technology is making it easier for criminals to impersonate executives, manipulate identity checks and launch sophisticated attacks at a scale that was previously difficult or expensive to achieve.

The implications are becoming harder to ignore. For banks, fintech companies, telecommunications operators and other businesses building their operations around digital infrastructure, the threat is no longer limited to stolen passwords or suspicious emails. Criminals can increasingly manufacture convincing identities, imitate familiar voices and exploit the trust employees place in digital communication.

That was a central concern at the 2026 annual cybersecurity conference of the Information Security Society of Africa–Nigeria (ISSAN), held in Lagos on Wednesday, November 18. Under the theme Cybersecurity: Balancing Risk, Resilience and Responsibility in the AI Era, industry professionals and policymakers examined the security challenges accompanying the rapid adoption of artificial intelligence.

The discussions took on added significance following a warning from the Central Bank of Nigeria (CBN) about the risks associated with giving AI systems excessive control over critical financial operations.

Director of the CBN’s Payments System Supervision Department, Rakiya Yusuf, cautioned financial institutions against allowing AI agents unrestricted access to customer information, institutional infrastructure and sensitive data.

"For safety and resilience, AI should never be given more access, authority, or autonomy than the implementing institutions can safely manage," she said.

Her warning highlights an important dimension of the emerging threat: businesses must protect themselves not only against criminals using AI, but also against vulnerabilities created by their own deployment of the technology. An AI system with excessive permissions, weak safeguards or inadequate human oversight could become another route into an organisation's most sensitive operations.

The scale of Africa's cybercrime problem

The warning comes against a deteriorating regional threat landscape. In its African Cyberthreat Assessment Report 2026, published on August 3, INTERPOL found that artificial intelligence was enabling 55 per cent of reported cybercrimes across Africa. The assessment drew on survey information from 36 African member countries.

The report also put cybercrime-related losses at $484 million, compared with $192 million in 2024, reflecting a sharp increase in the financial cost of digital attacks. Meanwhile, 72 per cent of surveyed countries reported the presence of scam centres.

INTERPOL's findings suggest that cybercrime is becoming increasingly organised and industrialised. Rather than relying on isolated individuals carrying out conventional online fraud, criminal networks are using digital tools to reach more victims, automate operations and exploit weaknesses across borders.

AI makes this easier. Generative tools can produce convincing phishing messages, imitate voices, manipulate videos and create fabricated identity documents. These capabilities can make fraudulent requests appear to come from trusted colleagues, business partners or public officials.

Consider a familiar corporate scenario: an employee receives a voice message that sounds exactly like the company's chief executive, urgently requesting a transfer to a new account. The voice is convincing, the request appears legitimate and the pressure to act quickly discourages verification. Without an independent approval process, a routine business transaction can become a costly fraud incident.

When a familiar face can no longer be trusted

The threat is not hypothetical. In Ghana, police arrested 11 suspects in May 2026 over allegations that they used AI-generated deepfake videos to impersonate President John Dramani Mahama. According to the Ghana News Agency, preliminary investigations indicated that the videos were used to solicit money and sensitive personal information from unsuspecting members of the public.

The arrests illustrate how AI-enabled impersonation can extend beyond corporate fraud into public-facing scams, exploiting the credibility attached to political leaders and other prominent personalities.

Identity verification is another growing vulnerability. A January 2025 report by Lagos-based identity verification company Smile ID found that deepfake incidents increased sevenfold between the second and fourth quarters of 2024. The company also reported that selfie anomalies accounted for 34 per cent of emerging biometric fraud cases.

The findings were based on more than 110 million identity verification checks conducted across Africa during 2024. They point to a growing challenge for businesses that depend on facial recognition and digital onboarding to establish whether customers are who they claim to be.

For fintech companies and digital lenders, the implications are particularly serious. Weak identity controls can allow criminals to create fraudulent accounts, take over legitimate accounts or exploit gaps in customer authentication.

Cyberattacks have a business cost beyond stolen data

AI is not responsible for every cyberattack, and two major corporate incidents demonstrate why the wider cybersecurity problem deserves attention. In April 2025, MTN Group disclosed that unauthorised parties had accessed personal information belonging to some customers in certain markets. The telecommunications group said its core network, billing systems and financial services infrastructure remained secure.

A month earlier, South African poultry producer Astral Foods reported a cyber incident that disrupted processing and customer deliveries. The company estimated that the resulting revenue losses and costs associated with clearing production backlogs reduced profits by approximately R20 million.

Neither company publicly attributed its incident to artificial intelligence. However, both illustrate the potential financial and operational consequences of inadequate cyber resilience.

For businesses operating in Africa's increasingly interconnected digital economy, the cost of an attack can extend beyond immediate financial losses. Service disruptions, regulatory investigations, legal expenses and damaged customer relationships can weaken a company's competitive position and undermine investor confidence.

The next challenge is protecting the systems businesses are building

The CBN's warning points to a broader responsibility for financial institutions as AI becomes more deeply embedded in banking and payment infrastructure. Businesses must ensure that the technology introduced to improve efficiency does not create new vulnerabilities.

That requires more than installing antivirus software or strengthening passwords. Organisations need stronger identity verification, multi-factor authentication, restricted access to sensitive systems, regular security testing and clear procedures for responding to suspicious activity.

For high-value transactions, voice messages, video calls and emails should never serve as the sole basis for authorisation. Independent verification through established approval channels is increasingly important, particularly where employees are asked to change payment details, transfer funds or disclose confidential information.

Businesses must also train employees to recognise phishing attempts, manipulated media and unusual requests, while ensuring that their incident-response plans account for attacks involving AI-generated content.

The challenge extends beyond individual companies. Financial institutions, regulators, telecommunications operators and technology providers share infrastructure and exchange data. A vulnerability in one part of that ecosystem can create risks for others, making information-sharing and coordinated incident response essential.

WHY IT MATTERS

Africa's digital economy depends on trust. Customers must believe that the person opening a bank account is genuine, that their money will reach the intended recipient and that the businesses holding their personal information can protect it. AI is putting that trust under pressure. For African businesses, cybersecurity must therefore become a boardroom responsibility rather than a problem delegated exclusively to IT departments. Investment in digital infrastructure without corresponding investment in security, governance and institutional capacity leaves businesses exposed to risks that can quickly translate into financial losses and reputational damage.

SOURCES & TRANSPARENCY
BusinessDay, INTERPOL — read the original source

Deepfake Attacks | Titas Gupta

Request a correction

Henry Ani

I'm a writer and a researcher. I love to see how the world changes and how much people in power controls that change. I'd love reporting on such issues