There is a woman I often think about when discussions turn to Nigeria’s new tax laws. She sells ewa agoyin from a small canteen in Ogba. She opened a bank account three years ago, not out of love for digital banking, but because receiving transfers was faster than waiting for customers to find change.
While I do not know every detail of how she runs her finances, I know enough about how millions of Nigerians like her operate to make a reasonable guess: that single account likely handles everything, business income, family support, school fees, and household contributions. There is no separation between “personal” and “business.” She has run her canteen for nearly a decade, and no one has ever explained that she might have tax obligations. The system never came looking for her either. That changed in 2025.
The federal government signed four interconnected tax reform laws last June, all taking effect January 01, 2026. In one legislative sweep, the government consolidated over sixty existing taxes and hundreds of unofficial levies into a unified framework. The reasoning is hard to argue with. Nigeria collects far less tax revenue relative to the size of its economy than almost any comparable country. Our tax-to-GDP ratio sits around 7.9%, the African average is roughly double that.
Part of the reason the ratio is so low is somewhat structural. About 93% of Nigeria’s employed population works in the informal economy, according to the National Bureau of Statistics nd they never had a reason to engage with the tax system. Only an estimated 15% to 20% of working-age Nigerians were captured by the tax system before these reforms. The new laws are designed to change that. Much of the design is thoughtful. Individuals earning below ₦800,000 annually owe no personal income tax. The stated goal is not to tax the poor harder but to introduce structure; every income-earning Nigerian with a National Identification Number now automatically has a Tax ID. Every registered company’s CAC number doubles as its Tax ID. The net has been cast very wide. The question now is, how do the people caught in it even begin to navigate it?
When the laws took effect in January, the first reaction was telling. Nigerians turned to bank transfer narrations, labelling inflows as “gift,” or “family support.” WhatsApp messages circulated with tips on what to write. The government had to clarify that the system relies on self-declaration, not automated surveillance of transaction descriptions. What this episode revealed was not tax evasion, but confusion. If educated, digitally savvy Nigerians panicked, what hope does the ewa agoyin seller have?
This is where generative AI enters the conversation, not as a cure-all or a buzzword but as a potential bridge. At its core, the compliance problem is personal, not theoretical. People need to know: what counts as income for me? What do I owe, if anything and how do I remit my tax? Today, getting that answer would require her to find the relevant sections of the Nigeria Tax Act, interpret the language, calculate her annual income, determine her tax bracket, subtract applicable reliefs, and then figure out how to file with the relevant revenue authority.
Generative AI could collapse that complexity into a conversation. It can read dense tax laws and explain them in plain language. It can analyse transaction patterns, amounts, frequency, timing to help individuals understand their own finances. With consent, it could guide users step by step through compliance, confirming exemptions or obligations.
Early signals already exist. Nigerian startups are experimenting with AI-assisted tax tools, and globally, similar models are emerging. The infrastructure is not hypothetical. Still, risks are real. Trust in automated systems is low. Data quality is uneven. Privacy laws require explicit consent. And accuracy matters; wrong advice in taxation has consequences.
Expecting Nigeria to go from widespread financial illiteracy to AI-powered tax compliance feels like skipping too many steps. Yet Nigeria has leapfrogged before. We skipped landlines for mobile phones. We moved from cash (most Nigerians have never held a chequebook) to mobile money at scale. Agency banking replaced bank branches where branches never existed. Not every leap landed perfectly, but the pattern holds. When formal infrastructure is absent and daily needs are pressing, Nigerians find a way to jump.
Nigeria’s new tax laws have created, perhaps unintentionally, the most compelling use case for generative AI in Nigerian finance that we have seen yet. Millions of people now have formal tax obligations for the first time. Most do not understand what those obligations are. They do not wish to evade tax, they simply exist in a system that never explained itself. Nigeria may offer a model for formalisation at scale. If not, the gap between policy and practice will widen. Could generative AI be the thing that bridges the gap?
