Artificial intelligence is becoming an increasingly familiar part of business conversations. From banks and online retailers to travel companies and telecommunications operators, organisations are exploring how automated systems can help them understand customers, set prices and make decisions faster.

But as the enthusiasm grows, so do questions about what happens when the information behind those decisions is inaccurate, the technology produces unexpected results, or nobody can explain why a particular decision was made.

Technology professional Oluwafemi Akinwande has previously raised some of these concerns in articles published by The Guardian, including her examination of pricing bias in African digital markets and the challenges of regulating artificial intelligence in Nigeria.

Her professional experience spans telecommunications, business analytics and digital product development, including work involving pricing intelligence and interconnected business systems. In this interview, she discusses why adopting AI requires more than investing in new technology.

There is so much excitement about artificial intelligence. Do you think businesses are moving too quickly?

I don’t think the problem is necessarily how quickly businesses are adopting AI. The bigger question is whether they understand what they are adopting and what they expect it to achieve.

Sometimes an organisation wants an AI solution before it has properly identified the problem. You still need to understand your processes, your data and the people who will use the technology. Otherwise, you may simply automate an existing problem.

You have written about pricing bias in digital markets. How does that concern relate to everyday business decisions?

Pricing is a good example because customers see the outcome, but they rarely know what happens behind it.

A price may be influenced by demand, availability, supplier information or other commercial factors. When automated systems are involved, businesses need ways to recognise unusual results and understand why they occurred.

Not every price difference is unfair, and not every unexpected result is caused by AI. The important thing is being able to investigate rather than assuming the system must be correct.

Your work has also involved pricing intelligence and digital platforms. What have those experiences taught you?

One thing I have learned is that technology problems are not always caused by the technology itself.

You may have different systems exchanging information, different teams responsible for different parts of a process, and business rules that are not properly understood by everyone involved.

Before introducing automation, organisations need to know where their information comes from, how it is used and what should happen when something goes wrong.

What about smaller African businesses that may not have large technology budgets?

They don’t have to begin with expensive AI systems. They can start by improving how they collect information, maintain records and make decisions.

If a business cannot trust its basic data, introducing a more sophisticated system will not automatically solve that problem.

It is better to identify a genuine business need and choose technology that addresses it.

You have also questioned whether Nigeria should simply adopt AI regulations developed elsewhere. Why?

Because the circumstances are not always the same.

Nigeria has its own business environment, infrastructure challenges and patterns of technology use. We can learn from international approaches, but regulation should also consider how technology is actually being used here.

The goal should be to encourage innovation while ensuring that businesses remain accountable for decisions affecting people.

What would you advise a company preparing to introduce AI into its operations?

Start with the problem, not the technology.

Understand your data, involve the people who know the business processes and decide how you will measure whether the system is working.

Most importantly, make sure someone remains responsible for the outcome. Automation should help people make better decisions, not become an excuse for decisions nobody can explain.