Artificial intelligence tools are giving cybercriminals skills they never had, and several recent reports show how fast the danger is growing.
Bad guys are using modified AI tools like "GhostGPT" to write code, build fake websites, and craft perfect scam messages without any technical background. The typos and bad grammar that used to give scammers away are gone. They use deepfake video software to trick bank security systems and run massive fake accounts. These tools scrape social media profiles to steal photos and voice clips, creating fake identities that can bypass standard identity checks.
Kenya, as usual, sits right in the middle of this digital storm. AI-assisted social engineering has been blamed for a wave of SIM-swap fraud that compromised more than 123,000 SIM cards and drained Sh491 million from mobile wallets and bank accounts.
Deepfake videos of well-known politicians have been circulating online, luring unsuspecting citizens into crypto scams. At the same time, hackers broke into President William Ruto's official website and demanded a Bitcoin payment.
Next door in Tanzania, fraudsters fabricated a deepfake of a government official issuing a false disaster warning during the Dar es Salaam floods, and the panic that followed was real enough.
In Uganda, criminals are pairing AI tools with stolen national ID details to hijack SIM lines and empty mobile money accounts before the owner even notices.
It all reads like a horror story, yet there is a stubborn pattern in how societies mature. Crime, terrorism and full-blown crises have a peculiar way of forcing nations to grow up faster than most 10-year development plans ever could.
In calm times, governments get comfortable and old problems fester quietly in some ministry drawer. It is pain, more often than planning documents, that builds better systems. When an institution stares down the barrel of its own collapse, the people running it have no room for excuses and must act.
Consider what followed the 1998 US Embassy bombings in Nairobi and Dar es Salaam. That single, horrific morning dragged East Africa's entire border security apparatus into the modern age almost overnight. Paper ledgers gave way to digital fingerprint scanners, scattered records were pulled into computerised intelligence databases, and border officers who had spent whole careers writing visitor names by hand were suddenly being retrained on new technology.
Before the blasts, immigration desks still ran on ink and handwriting. It took that scale of horror to make the State discard the exercise books and wire border posts in favour of central systems.
Roughly a decade later, the 2007 post-election violence closed major highways, torched shops and shut down bank branches across the country. Families needed a way to reach relatives trapped behind roadblocks, and hand-delivered cash was no longer an option.
That desperation is what pushed a fledgling two-year-old M-Pesa from a modest experiment in sharing airtime between friends into the global mobile money backbone it is today. Once branches were burning and roads were cut by ethnic militias, moving money by text stopped being a novelty and became survival.
The Westgate Mall and Garissa University attacks forced a similar reckoning. Kenya built a central cyber intelligence unit, assembled digital threat-response teams, and deployed thousands of smart cameras with facial and number-plate recognition across Nairobi and Mombasa.
Those attacks exposed just how badly the different security agencies had been talking past one another. That failure is what finally pushed police, military and digital intelligence into one connected network, with specialised teams now watching for online radicalisation.
Now a stranger threat is emerging. Boko Haram in Nigeria and various jihadist outfits scattered across the Sahel have discovered Elon Musk's Starlink. They are running entire operations from deep inside forests and deserts that no cell tower ever reached.
They stream propaganda, coordinate ambushes and run networks beyond the reach of any telecom operator a government could pressure.
In Sudan, armed factions kept Starlink dishes running despite the country's phone networks going dark. Fighters now carry small satellite terminals into remote hideouts, making encrypted calls, tracking troop movements, and broadcasting straight to the internet without ever touching the local networks that police would normally monitor.
That is forcing governments and banks to catch up quickly. Militaries are creating satellite-tracking units, and African banks are racing to deploy AI systems to catch fraud in real time. Security forces across West Africa and the Horn are forming spectrum-monitoring teams to spot unauthorised satellite signals.
At the same time, it's been reported that major banks across Kenya and East Africa are rolling out machine-learning models that watch customer behaviour, flag takeovers within milliseconds, and freeze stolen funds before they can be cashed out.
Left to their own devices, without these threats breathing down their necks, the State and the private sector would probably have stayed comfortable, clinging to old systems and endlessly postponing the hard investments. But that is how it works here: every time criminals get cleverer, the rest of society is dragged, kicking and complaining, into getting cleverer too.
Each fresh nightmare, however unwelcome, quietly becomes the engine for the next leap forward.
The author is a journalist, writer and curator of the Wall of Great Africans. X@cobbo3
