If you’re a Safaricom user in Kenya right now, your SMS inbox likely looks like a crime scene. Between unsolicited betting odds, predatory loan offers, and “promotional” messages you never signed up for, the noise is deafening. The situation has gotten so bad that the Consumers Federation of Kenya (COFEK) recently raised a red flag, citing a massive surge in spam that is pushing data privacy concerns to the brink.
But here’s the thing: I’m not seeing any of it.
On my Airtel line, my inbox is a ghost town—in the best way possible. Aside from the occasional official update from Airtel themselves, I receive zero spam. No loan offers, betting tips, “quick cash” adverts, shady investment pitches, nothing. It begs the question: What gives? Is Safaricom carelessly leaking data? Is Airtel’s new AI tech effectively bulletproof? Or is there a simpler, more economic reason why scammers seem to be ignoring the second-largest telco in the country?
Is Safaricom a victim of success?
First, let’s address the elephant in the room. Safaricom is the undeniable king of the Kenyan mobile market. With M-PESA being the lifeblood of the economy, it is naturally the most lucrative target for spammers. If you are a betting firm or a shady loan shark, you want to target users who transact frequently and have mobile money active. That demographic is overwhelmingly on Safaricom. It’s highly probable that these “marketing” firms are simply playing the numbers game, buying databases that are heavily skewed towards the market leader.
But that doesn’t let Safaricom off the hook. The sheer volume of complaints suggests a systemic failure to police bulk SMS providers who abuse their API access to flood users with junk.
AI spam alert is Airtel’s secret weapon
While Safaricom users struggle to find the “Stop” keyword, Airtel has been quietly deploying what might be the most sophisticated anti-spam tool in Africa. Earlier this year, the telco launched its AI Spam Alert Service, a solution they bill as a first for the continent. Unlike traditional blockers that rely on simple keyword filtering, this system is surprisingly complex and operates entirely in the background.
What makes this feature particularly interesting is how it handles privacy. The Artificial Intelligence does not actually read the content of your individual SMS messages. Instead, it analyzes sender behavior in real-time across over 250 distinct parameters. It looks for specific “bad actor” patterns, such as how often a sender changes their SIM card or the volume of people they are messaging simultaneously. It also analyzes the geographical spread of the targets, for instance, flagging a number that is blasting messages across the entire map of Kenya rather than a localized area. It even checks the “send-to-receive” ratio to see if the user only ever pushes messages out without ever receiving replies, a classic hallmark of a spammer.
The speed at which this happens is staggering. Airtel claims the AI completes this 250-parameter check in just 2 milliseconds, literally faster than the time it takes for you to blink your eye.
Perhaps the best part of this implementation is its accessibility. Airtel has made the service completely free and automatically activated for every single subscriber. There are no applications to install, no settings to toggle, and no subscriptions to manage. Crucially, it is device agnostic, meaning it works just as well for a grandmother using a feature phone (“mulika mwizi”) as it does for a tech enthusiast on the latest flagship smartphone.
Here is where my experience gets interesting.
I have never actually seen a “Suspected SPAM” warning on my Airtel line, which is how the system is designed to alert you when a dubious message slips through. Does that mean the feature isn’t working? Not exactly.
On the contrary, I suspect it might be working so well that the spam isn’t even reaching the “suspicion” stage. However, the more likely reality is a mix of two factors. First, Airtel’s network-level blocking might be nuking highly malicious spam before it even tries to ring my phone. Second, spammers simply aren’t wasting their bulk SMS credits on Airtel numbers yet because the ROI is lower than targeting Safaricom’s massive base.
The contrast between the two networks right now is stark. On one side, you have Safaricom users petitioning regulators because their phones feel like public billboards. On the other, Airtel users are enjoying a quiet, premium experience that is, ironically, for free.
If Airtel’s AI is indeed the primary reason for this silence, Safaricom needs to take notes immediately. Relying on users to forward spam to ‘333’ is archaic. We are in the era of AI; the network should be doing the heavy lifting, not the customer. For now, I’ll keep enjoying the silence on my Airtel line. But I can’t help but wonder: if the spammers ever decide to shift their focus, will the AI shield hold up? Judging by the specs, Airtel seems more than ready for the fight. Safaricom, it’s your move.
For our readers, what’s your story? Do you get spam SMS and on which network? Drop your experience in the comments.
