Africa’s reported cybercrime losses didn’t drift higher last year. They jumped from $192 million to $484 million, more than doubling, and Interpol’s latest report names a specific culprit for the surge: AI is now present in 55 percent of cybercrimes reported on the continent.
That figure deserves a pause. It isn’t a forecast, and it isn’t a poll of what security teams expect to face. It’s a proportion of cases that have already been filed, compiled from data spanning 36 African countries.
2025: the year the tool started running the show
Interpol treats 2025 as a dividing line. Up to that point, AI was something criminals reached for to move quicker. Past it, the agency describes AI as the “core operational driver of cybercrime in Africa” on the continent.
That shift in language carries weight. A handy tool makes an existing crew faster. An operational driver redefines what a crew that size is even capable of taking on.
“AI is automating every stage of a cyberattack from reconnaissance and phishing to extortion and evasion,” said Neal Jetton, head of Interpol’s cybercrime unit.
Deepfake extortion on a scale that’s hard to picture
Roughly 600,000 cases of digital extortion involving deepfakes were logged. That number recasts the rest of the report, because extortion happening at that volume isn’t a targeted campaign. It’s throughput.
On the fraud side, synthetic identities are pulling similar duty. According to the report, they’re being produced convincingly enough to slip past biometric systems — the detail that ought to unsettle anyone who has spent recent years hearing that a face scan or voice print is the dependable successor to the password.
By Jetton’s account, no part of the attack chain is untouched. Reconnaissance. Phishing. Extortion. Evasion. There isn’t one weak point to reinforce, which explains why the loss total moved the way it did.
Enforcement is scoring wins, but look at the ratio
The report isn’t written as pure alarm. Four Interpol-coordinated operations spanning 2025 and 2026 delivered more than 1,500 arrests along with seizures topping $100 million.
That’s genuine progress, and $100 million recovered is hardly a rounding error. Weigh it against $484 million in reported losses over the same stretch, though, and the arithmetic makes the position clear. Enforcement is recovering a slice while the losses keep stacking up.
And $484 million captures only what was actually reported. Anyone who follows fraud statistics knows reported losses lag the real total, often badly, and that gap tends to stretch wider when attacks are automated and each one is small on its own.
What the numbers are telling you
If your security setup anywhere treats biometrics as settled, the synthetic identity finding is the point to act on. Interpol is documenting systems being defeated right now, in cases already tallied — not a proof of concept in a lab.
The 55 percent number is the one to keep watching. It’s the metric that distinguishes a rough year from a structural change, and data from 36 countries provides a wide enough base to see it shift.
