Sunday, November 22, 2026

Subscribe

NewsEdition KenyaKenya

AI industrialises cybercrime in Africa, outpacing investigators

Deepfakes, sextortion and synthetic identities are fuelling fraud in Africa. INTERPOL also warns that law enforcement remains undertrained to tackle these automated attacks.

Bright polling station in Maputo with anonymous voters seen from behind, a transparent ballot box and ballots with no legible text

When fraud starts speaking and writing for itself

Cyberfraud in Africa is changing shape. It no longer depends solely on skilled technicians or highly organised networks. With readily available AI tools, scammers can now write convincing messages, clone a voice, create a video and tailor their approach to their target. This industrialisation of scams directly affects individuals, small businesses and institutions that are going digital faster than their defences can keep up.

INTERPOL’s 2025 Africa cyberthreat assessment report describes an environment in which the growth of mobile services, online commerce and digital payments is also opening new doors to criminals. The organisation says that more than two-thirds of the African countries surveyed consider cybercrime to account for a medium to high share of all offences, and that in both West and East Africa, more than 30% of reported crimes already take a digital form.

What the report says, and what to take from it

In this continent-wide snapshot, AI is no longer a mere accessory. INTERPOL says that 55% of cybercrime cases recorded in 2025 involved AI in some form, occasionally or frequently. The same document also explains that cybercriminals now use AI throughout the attack cycle: identifying victims, targeted phishing, extortion, and then evading detection systems.

The finding is not limited to a theoretical threat. The report cites fraud fuelled by manipulated content, including doctored videos promoting fake financial investments. It also notes a rise in sextortion scams — extortion using real or fabricated sexual images — and says synthetic or altered images are increasingly being used to put pressure on victims.

INTERPOL is not talking about a distant future. The organisation is already seeing practical uses of cloned voices, doctored videos and AI-generated emails to make attacks more credible. It also warns that African countries remain unevenly prepared for these practices, with AI adoption among law enforcement agencies still limited.

Deepfakes, sextortion, synthetic identities: the attack chain grows

Deepfakes have become one of the most visible symbols of this shift. INTERPOL explains that manipulated content is used to impersonate public officials, business executives or influential figures to induce fraudulent investments or extract money. The report also links this trend to a rise in so-called business email compromise attacks, in which an attacker poses as a legitimate contact to obtain a bank transfer or sensitive information.

AI is also taking sextortion to a new scale. INTERPOL’s report notes that sexual impersonation and the creation of explicit images can now be done using a single public photograph. Those most at risk are often young and highly connected, but the phenomenon also affects adults, women, men, students and workers in the informal sector. The issue is therefore not only a policing matter; it is social, psychological and economic.

Another worrying development is the creation of synthetic identities. The report says these profiles, combining real data with invented details, are used to open bank accounts, obtain mobile loans or register SIM cards under false identities. This method undermines KYC checks — the customer identity verification procedures used by banks and telecoms operators. INTERPOL cites cases in Cameroon, Mali and Tanzania in particular.

Police move more slowly than scammers

The real imbalance lies in the public response. According to INTERPOL, only 33% of the agencies surveyed use AI to detect threats, and 31% use it for digital forensics. Just 8% of intelligence analysts have advanced AI expertise, while 92% of agencies cite a lack of technical skills as their main obstacle. The problem is therefore not simply the purchase of software; it is also human, organisational and financial.

This delay weighs on investigations on the ground. A cloned voice call, fake video or highly personalised phishing message can still deceive officers who are not well trained in new uses of AI. In several countries, specialist units exist, but they lack equipment, access to data and swift cooperation with telecoms operators, banks and digital platforms. The consequence is clear: the more professional fraud becomes, the more institutions fighting it must move beyond a purely manual approach.

This vulnerability is not uniform. South Africa, Kenya, Nigeria and Ghana already have more developed cybersecurity ecosystems than some other markets, but pressure remains high in areas where mobile use is surging and training is progressing more slowly than digitisation. INTERPOL’s report also stresses the differences in capacity between regions and the importance of faster cross-border cooperation.

A continental challenge that goes beyond policing

The financial losses give a sense of the scale of the problem. INTERPOL estimates that cyber incidents caused more than $3 billion in losses across the continent between 2019 and 2025. The organisation also says that between 2024 and 2025, losses linked to cybercrime rose from $192 million to $484 million, while the number of identified victims increased from 35,000 to 87,000. These figures are consolidated estimates, but they show a clear trend: digital fraud is costing households, businesses and states more and more.

The issue also has a regional dimension. Operations coordinated by INTERPOL in Africa, such as those carried out with national services under the AFJOC framework, show that criminal networks move from one country to another much faster than judicial procedures do. Cybersecurity is therefore a matter of practical sovereignty as well as a technical challenge. It affects banks, telecoms, tax authorities, hospitals and payment platforms.

The next steps will involve two fronts. On the one hand, states will need to train investigators, magistrates and regulators to recognise AI-generated content and trace the financial flows that fuel these frauds. On the other, regional and continental organisations will need to speed up information-sharing, because an attack launched remotely can hit several countries within hours. That is where Africa’s advantage will lie: not in a simple race to acquire tools, but in the ability to turn cooperation into an operational reflex.

Automatic translation. Original French version

News quiz

Edition KenyaKenya: the news quiz for 12–13 November

5 questions