As AI agents become increasingly capable of handling tasks on behalf of users, fintech companies are beginning to explore how AI systems can move beyond answering questions and into taking actions, including making purchases, booking services, and managing financial transactions.
Paystack is now bringing that idea into payments with the launch of Paystack Index, an early-access product developed by Paystack with support from TSG Labs.
The launch is the first public product to emerge from TSG Labs since Paystack announced the creation of The Stack Group in January 2026. The holding company brings together Paystack, Zap, Paystack Microfinance Bank, and TSG Labs, a venture studio focused on building experimental products that may be too early or uncertain for the group’s more established businesses.
“If you take a step back and compare the number of people that are using their browser for commerce and now introducing agents for discovery, you’ll see that agents are coming up and browsers are going down,” Paystack CEO Shola Akinlade says.
The company sees Paystack Index as an early experiment in extending its payment infrastructure into AI-powered experiences. Rather than navigating websites and apps manually, users can ask supported AI agents to complete transactions on their behalf. Paystack Index interprets the request, routes it to the appropriate merchant or service provider, processes the transaction through Paystack and Zap’s infrastructure, and helps complete checkout within the AI experience.
The product is currently live in Nigeria and works with AI clients, including ChatGPT, Claude, and OpenClaw. At launch, supported use cases include purchasing airtime and mobile data, funding wallets, sending money through Zap, and ordering food through Chowdeck.
Currently, Paystack is not positioning the product around complex purchasing decisions. The initial use cases are largely routine and repeatable transactions that users already perform frequently. Food ordering is perhaps the most complex category currently supported, but even that follows a relatively predictable purchase flow.
Paystack is also maintaining tight control over the experience. The company currently curates merchants, though Akinlade says that approach could evolve as usage grows and the company learns more about how customers and businesses interact through AI agents.
Building checkout for an AI-first internet
The launch represents an early bet on agentic commerce. While enthusiasm around AI agents has grown rapidly over the last year, Akinlade does not expect immediate mass adoption.
Victoria Fakiya – Senior Writer
Techpoint Digest
Stop struggling to find your tech career path
Discover in-demand tech skills and build a standout portfolio in this FREE 5-day email course
“The thing we know is that there’s a bunch of early adopters in Nigeria or in Africa. There are people that used ChatGPT on the first day and back to the spirit of TSG Labs, we wanted to start building with those early adopters,” he says.
For now, Paystack is focused on learning how users behave when AI agents become a purchasing interface and what infrastructure merchants need to participate safely. The company expects that process to take time and does not anticipate significant traction for at least another year.
The idea is also beginning to attract attention across Africa’s fintech ecosystem. Earlier this month, Anchor launched an MCP server that allows AI agents to access and interact with its API resources, signalling growing interest in building financial infrastructure for AI-native applications.
One of the questions raised by agentic commerce is what happens to the relationship between merchants and customers when an AI intermediary sits between them. If consumers increasingly rely on AI systems to discover products, compare options, and complete purchases, merchants could lose some of the direct interactions that traditionally helped build customer relationships.
Akinlade argues that this concern is not entirely new. In his view, a similar shift occurred when businesses moved from physical storefronts to eCommerce.
The transition from brick-and-mortar retail to online shopping removed many of the face-to-face interactions that once defined commerce. Businesses adapted by finding new ways to communicate their value online, whether through product pages, customer reviews, digital marketing, or personalised experiences. Agent-driven commerce may represent a similar evolution rather than a complete disruption.
In fact, Akinlade believes AI agents could ultimately strengthen how merchants communicate with customers.
“If you have a commerce website, you can’t talk to the website, but the way agentic commerce works, the merchant can actually add a lot of context to their website,” he says.
Safety and user control are likely to be critical factors in determining whether consumers become comfortable with AI agents participating in financial transactions. Concerns around agents acting independently or making unauthorised purchases have become common as the technology advances.
Paystack says users remain in control of transactions completed through Index. The system only acts on requests submitted through a user’s chosen AI agent, within the permissions and spending limits the user sets. Users can configure controls after signing up, creating guardrails around what an AI agent is authorised to do on their behalf.
For Paystack, the launch is less about capturing immediate transaction volume and more about understanding where commerce is headed. If AI agents do become a primary interface for discovering products and completing transactions, the company wants to be building the payment infrastructure for that future today rather than waiting for the market to mature.
