Artificial intelligence is no longer just making cybercrime more sophisticated in Africa. It is changing how cheaply and quickly criminals can operate. A new INTERPOL African Cyberthreat Assessment Report says AI is now linked to 55% of reported cybercrime across Africa, while reported financial losses reached about $484 million in 2025, up sharply from $192 million in 2024. The findings suggest that the continent is moving into a period where traditional scams can be industrialized, replicated and targeted at far greater scale.
The important part of the report is not simply the percentage. AI is giving criminals access to capabilities that previously required technical expertise, larger teams or considerable time.
Fraudsters can use AI to generate convincing phishing messages, impersonate people through synthetic voices and images, create fake identities and automate social engineering campaigns. That means the weakest point in many attacks is increasingly the victim’s ability to recognise something as fake, rather than the criminal’s ability to produce the deception.
Africa was already dealing with a serious cybercrime problem before generative AI became widely available. INTERPOL’s previous assessment found that cyber-related offences accounted for more than 30% of reported crime in Western and Eastern Africa, while 90% of African countries surveyed said their law-enforcement or prosecution capacity needed significant improvement.
Online scams, business email compromise, ransomware and digital sextortion were already among the major threats. AI is now sitting on top of that existing infrastructure and making many of those crimes easier to scale.
That creates a particularly difficult problem for countries such as Nigeria, where digital payments, mobile banking and social media have become deeply embedded in everyday life. The more economic activity moves online, the more valuable those systems become to organised criminals.
A convincing AI-generated voice pretending to be a relative, a fake customer-service account or a highly personalised investment pitch can reach thousands of potential victims without a scammer having to manually write every message. The threat is therefore moving from individual fraudsters towards something closer to cybercrime businesses with automated production lines.
The defensive response will have to change accordingly. Arresting scammers after money has disappeared is increasingly inadequate when criminals can quickly replace accounts, identities and infrastructure.
INTERPOL has already demonstrated the value of coordinated operations, with a 2026 operation involving 16 African countries resulting in 651 arrests and the recovery of more than $4.3 million. But enforcement will also need stronger real-time information sharing between banks, telecommunications companies, technology platforms and law-enforcement agencies.
There is a more uncomfortable implication for African businesses and consumers. The old security habit of checking whether an email “looks suspicious” is becoming less useful when AI can produce fluent, personalised and context-aware communication. Financial institutions will increasingly need systems that verify behaviour and identity rather than relying primarily on what a customer says or what a message looks like.
AI is therefore creating a race between attackers and defenders. Africa cannot afford to enter that race with yesterday’s cybersecurity tools. If criminals are using automation to make fraud faster, cheaper and more convincing, governments and financial institutions will need to use automation and AI defensively at the same scale.
