As Nigeria grapples with persistent challenges of budget padding, misappropriation, and inefficiencies in public finance management, the role of AI in promoting transparency has come under renewed focus.
In this exclusive interview, Dr. Nasir Aminu, a Senior Lecturer in Economics and Finance and a Senior Fellow of Advanced Higher Education at Cardiff Metropolitan University, sheds light on how artificial intelligence (AI) can be a game-changer in detecting fraud, improving fiscal discipline, and enhancing overall budget efficiency.
Drawing from global best practices and his deep understanding of Nigeria’s economic landscape, Dr. Aminu makes a compelling case for the strategic deployment of AI in government budgeting processes.
“So, there is room for it, especially in Nigeria, where we have a culture of slow and lack of inefficient work. GenAI can enhance and make things more efficient,” Dr. Aminu noted.
Naijaonpoint: Let’s begin with your most recent research, conducted in collaboration with scholars from the United Kingdom and Bulgaria, which explores the relevance of generative artificial intelligence chatbots in investment decision-making. Could you give us an overview of the study?
Dr. Nasir: ChatGPT came out, and it was being used across sectors. But what we should understand is: people are not very conversant with finance and investment decisions. And we realized that there should be a literature gap to fill in that area to see how accurate. Because it has been discussed all across the news media that people ask AI, people do not want to go to a broker, and so on, because of the charges. So, they may end up resorting to, instead of Googling, a chatbot like ChatGPT, Microsoft Bing, or Gemini to ask for investment or finance decisions.
We tried to take a very small sample of the UK, which is a very strong financial centre – some call it the centre of finance in the world, even though it doesn’t have the capital of a country like America. And we went on to another lower-income developed economy in Europe, which is Bulgaria. This was done based on purposive sampling because we have access to that country, and partnerships and research partners are from there. We went on to combine both and said, ‘Let’s do this.’
Our method was a two-stage mixed approach where we conducted a quantitative questionnaire, and we were asking for accuracy and how specific the chatbots were.
We then asked individuals – a financial expert – who either teaches finance or an investor to verify the outputs from the chatbots.
What did we do? We’ve already asked ChatGPT, for instance, “We have 50 Pounds to invest today. Where will you recommend me to invest in the UK?” and we did the same thing on the same day in Bulgaria.
After that, we went further to ask, ‘We have 1000 Pounds today. Where and how do you suggest we invest in the UK stock exchange market?’ We repeated the same for Bulgaria.
While we did that simultaneously, there was a translation for Bulgaria because we do not want ChatGPT to learn through machine learning, so that it doesn’t store that data.
After a week, we came back to doing the same thing. When we did the same thing, it gave us different information.
We had about ten responses and asked financial experts who are teachers or investors to rate them as to whether they are relevant, accurate, or how specific and the justification for what ChatGPT is. We concluded that the prompts that we provided did not significantly influence the responses in Bulgaria but was significant for UK investments.
The originality of the paper is that it assesses the relevance, accuracy, specificity, and justification of GenAI chatbots’ investment recommendations for two different periods, investment amounts, and countries.
Naijaonpoint: How do you see this study being replicated or adapted for the Nigerian business environment?
Naijaonpoint: How do you see this study being replicated or adapted for the Nigerian business environment?
Dr. Nasir: Yes, absolutely. These are things that can be replicated not just in Nigeria but in emerging economies as well as in fully developed economies. This all depends on the level of education of investors. It also depends on the language that investors pick.
For example, Nigerians rarely invest in insurance companies, you know? One of the problems is that we have growing insurance companies. For this, you may see some skewness when it comes to that. And that is something that you want to understand.
Naijaonpoint: Considering the peculiarities of Nigeria’s financial ecosystem—such as regulatory instability, limited data, and FX volatility—what modifications would be necessary for GenAI to produce relevant and accurate investment advice locally?
Dr. Nasir: What AI does is more or less like an assistant. It’s a computer, basically. It only gives you the information it already has within its scope. Without the necessary data, it will be handicapped.
Regardless of whatever you do, humans, the final decision comes into play. You need that human touch. But let’s be positive about it. It does the work of assistants. It can do the spreadsheets for you. It can extrapolate data for you. Generative AI has been around for a while, even other coding systems.
Yes, it’s a very useful tool for day-to-day activities. It’s not just Nigeria even developed world like UK, US, etc. UK is using it for its healthcare system and going to be using it in its financial system to detect fraud, which Nigeria is suffering from.
A very good example that we can use generative AI for is the budget. Generative AI can detect fraud in our budgets. For example, you have 2100 lines and 285 pages in the 2025 budget. And those pages are all expenditures that are supposed to occur. So, GenAI will quickly, within a minute, detect padded elements in the line items.
There is this NGO, BudgIT, it took them about three months to come up with the loopholes. And those were just very few loopholes they were able to find, though some were debunked. But, GenAI could have done that for them within a few minutes.
When the budget was submitted, it would have helped members of Parliament or the National Assembly who are scrutinizing the budget to see that there were duplications or inconsistencies. The main thing is we want fiscal efficiency, which is what the World Bank and IMF have been talking about.
This is what makes you worry that you are spending over N300 billion on streetlights and boreholes. What happens to getting sustainable solutions to water and light? So, GenAI could have given them alternative solutions as quickly as possible, which could be theoretical. And if you have the data, they could feed that data to see what a collective decision would have looked like.
The padding they were talking about, GenAI can curb it. I was listening to one of the House members giving justification on why they are putting boreholes in the healthcare budget, or in places they are not supposed to be. With GenAI, it could have given them a balanced, more efficient approach to show them what they could have done without creating suspicion or leading to inefficiency.
One of the reasons we’ve not been able to implement the 2025 budget is because we haven’t finished the capital expenditure of 2024. There would have been a performance index that GenAI would have been able to measure and produce as quickly as possible.
So, there is room for it especially in Nigeria where we have a culture of slow and lack of efficient work. GenAI can enhance and make things more efficient.
Naijaonpoint: Based on your research, do you see GenAI replacing traditional investment advisors anytime soon—particularly in Nigeria?
Dr. Nasir: You can’t beat experience. GenAI has a machine learning facility there. But the human touch – you can’t beat it. GenAI is an additional tool for brokers. The stored information GenAI has will enhance brokers much more efficiently than if they don’t have GenAI.
But human decisions can factor in all the risks, but GenAI cannot. Relying on GenAI for financial decisions can be misleading. But it’s very important in our lives now.
Naijaonpoint: How would you assess the Tinubu administration’s efforts so far in attracting Foreign Direct Investment into Nigeria?
Dr. Nasir: The data speaks for itself. Nigeria’s FDI is about $1.08 billion compared to Egypt, which is about $14 billion. That tells you everything about it. And this is in Africa, comparatively. Kenya is a very small economy, but it’s having much more FDI. Why is that? Because the Nigerian economy is not attracting FDI.
I can go back in history and look at what happened a few years ago. The data shows that Nigeria is underperforming than many other African countries compared to what it used to be.
Tinubu’s administration? Yes, it should be a decent economy. But why are we having slower FDI? We have this because of how the economy is being run. How conducive is it for businesses? If foreign businesses cannot project profitability, then they are going to go somewhere else.
The Minister of Steel recently said Tinubu attracted over $500 million in FDI to the steel sector in the last 20 months. You will think that’s a very good thing, but look at what we are talking about since he’s been in power for over 25 months.
Comparatively, if you are going to look at [President Umar] Yar’Adua in 2007 alone, he generated $6 billion. In 2008, he generated $8.2 billion in FDI. The same with Goodluck Jonathan in 2011, despite the insecurity and national crisis, Nigeria generated $8 billion. Comparatively, if you look at Buhari’s first two years, he generated $3 billion.
This is Nigeria at its lowest. Why is this happening? It’s because of the policies that are ongoing.
Naijaonpoint: There are people who have started saying that instead of pursuing FDIs, Nigeria can export private capital through the likes of Dangote. What’s your take?