Watch my work ↗
AI / Commerce / Africa6 MIN READ

AI Agents Are About to Start Buying Things. Africa Might Leapfrog Again.

Card rails break when the buyer is a machine. The payment layer being built for AI agents runs on stablecoins — rails Africa already has a head start on.

By Victor MasokeFounder & builder
AI agents buying things and Africa's leapfrog moment

Picture this: you tell an AI agent, "find me waterproof hiking boots under $150 that arrive by Friday," and it searches multiple retailers, compares prices, checks reviews, verifies availability, and completes the purchase — payment and all — without you touching a checkout page. This isn't a demo of a distant future. It's happening now, it has a name — agentic commerce — and it's one of the fastest-moving stories in technology in 2026. Morgan Stanley projects hundreds of billions of dollars of US e-commerce will move to agentic channels by 2030.

Here's the part almost nobody is talking about: this shift may hand Africa another leapfrog moment, the way mobile money let the continent skip the era of bank branches entirely. To see why, you have to understand what breaks when a machine, not a human, becomes the one paying.

What "agentic commerce" actually means

Strip away the hype and agentic commerce is simple: an AI agent that doesn't just recommend a purchase but executes it — researching, deciding, and paying on your behalf, with authority you've delegated to it. Traditional automation (like autopay) follows fixed rules. An agent makes decisions to achieve a goal you set in plain language.

That distinction matters enormously at the moment of payment, because our entire payment system was built on one deep assumption: that a human is the one paying. And that assumption breaks in ways that turn out to be very interesting for Africa.

Why card rails break when the buyer is a machine

Credit-card infrastructure was designed for human-speed, human-scale transactions. Point an army of AI agents at it, transacting continuously, and it strains in specific ways:

Why Card Rails Strain Under Machine Buyers

flowchart TD M[AI agents transacting
continuously, at machine speed] --> A[Micropayments
fractions of a cent] M --> B[Thousands of txns/second
24/7, never sleeps] A --> A2[Card interchange fees
dwarf the payment] B --> B2[Chargebacks assume
a human dispute window] B --> B3[Settlement runs on
banking hours] A2 --> S[New rail needed:
stablecoins] B2 --> S B3 --> S S --> S2[Instant, 24/7, sub-cent,
no interchange, no chargeback]

The economics collapse at small sizes. Card interchange fees are built for purchases of dollars, not cents. When agents transact in micropayments — paying fractions of a cent for a piece of data, an API call, a sliver of compute — card fees make the transaction uneconomical. The fee dwarfs the payment.

Chargebacks assume a human dispute. The whole card dispute system is built around a person noticing a bad charge and contesting it within a window. That model doesn't map cleanly onto autonomous agents transacting thousands of times a second.

Settlement runs on banking hours. Agents operate 24/7, at machine speed. Traditional settlement — batched, delayed, business-hours-bound — is a mismatch for a buyer that never sleeps.

So a new payment layer is being built specifically for machine buyers. And the striking thing is what it's being built on: increasingly, stablecoins. Coinbase launched a platform in April 2026 where autonomous agents pay each other in stablecoins via a protocol called x402 — and within weeks it had tens of thousands of active agents and over a hundred million transactions, with an average value around 30 cents. Stablecoins settle instantly, 24/7, in fractions of a cent, with no interchange fee and no chargeback assumption. They are, almost by accident, a near-perfect fit for how machines want to pay.

Where Africa comes in

Now connect two trends I keep coming back to. First: Africa has already gone further on stablecoins than almost anywhere on earth — they account for a huge share of the continent's crypto volume, used not for speculation but for real payments, remittances, and dollar access. Second: agentic commerce runs best on exactly those stablecoin rails.

That's the leapfrog setup. The developed world is trying to bolt machine-speed, stablecoin-based payments onto a deeply entrenched card system that resists it — every incumbent, every bank, every card network has a reason to slow the transition. Africa doesn't carry that baggage in the same way. In many African contexts, there's no dominant card infrastructure to protect, mobile money already normalised phone-based payments, and stablecoins are already in serious use. The rails that agentic commerce wants are, in some ways, closer to what Africa already has than to what the West is trying to retrofit.

We've seen this movie before. Africa didn't build a continent of bank branches and then digitise them — it skipped straight to mobile money. It didn't wire millions of landlines — it went straight to mobile phones. The pattern is that the absence of entrenched legacy infrastructure, usually a disadvantage, becomes an advantage when a genuinely new paradigm arrives, because there's less to unlearn and less incumbency to fight. Agentic, stablecoin-based commerce might be the next instance of exactly that pattern.

The honest caveats

I want to be careful not to oversell this, because the same reality checks that apply everywhere apply here too.

Trust is the big one. Consumers everywhere — Africa included — are still wary of letting an AI spend their money autonomously. Adoption of fully autonomous purchasing will be gradual, starting with low-risk, low-value transactions and human approval for anything significant. The infrastructure for identity, authorization, and security in an agent-driven world is still immature. And Africa's real constraints — connectivity gaps, the coverage and trust issues I've written about before — don't vanish because a new paradigm arrived.

So this isn't "Africa wins agentic commerce automatically." It's that the structural fit between where Africa's payment behaviour already is and where machine payments are heading is unusually good — and that's an opening worth understanding early.

What to take from this

If you're building in African fintech or commerce, here's the signal. A new layer of payment infrastructure is being built for a buyer that isn't human, and it's being built on rails — instant, stablecoin-based, micro-denominated — that Africa has an unusual head start on. The infrastructure window for this is open right now, in 2026, before the standards and the winners are settled.

The builders who understand that agentic commerce is coming, that it breaks the card model, and that stablecoins are its native currency — and who realise Africa is structurally well-placed for exactly this transition — are looking at the same kind of opening the mobile-money pioneers had fifteen years ago. Most people will dismiss AI agents buying things as a rich-world novelty. The ones paying attention will notice it rhymes, precisely, with the leapfrog Africa has already pulled off once. The question isn't whether machines will start paying. It's who builds the rails they pay on — and why those rails might just be laid fastest in the place with the least to unlearn.


References

  1. Paz.ai — Agentic Commerce in 2026: How AI Agents Buy Products
  2. Eco — What Is Agentic Commerce? The 2026 Guide
  3. IMF — How Agentic AI Will Reshape Payments
  4. Mastercard — Mastercard launches Agent Pay for Machines
  5. PaySpace Magazine — Agentic Payments 2026
THE CONVERSATION CONTINUES

Come for the ideas.
Leave with clarity.

▶ Watch My Work

Tech, business, and the things I learn building both.