Today I have Nancy Gleason, Professor of Practice in Political Science at the Mohammed bin Zayed University of Artificial Intelligence. Her research focuses on the Fourth Industrial Revolution and the future of work. Nancy, thank you so much for joining us today.
Pleasure to be here.
Beyond chatbots — where is generative AI transforming the financial system in a meaningful way?
The real transformation is not in the chatbots. I always say ChatGPT is not the point. It's in the redesign of the financial institutions behind the chatbot. For JP Morgan, that's AI applied to fraud and risk. For Mastercard and Visa, it's real-time fraud detection. AI is increasingly embedded in anti-money laundering, credit underwriting, and insurance. The Bank of International Settlements identifies fraud detection and AML as among the most important current applications of AI. The meaningful shift is that AI can help institutions detect, decide, and document — but it has to be with humans in the loop.
Are we seeing real productivity gains from AI in finance or is this just higher valuations?
I think we are seeing real gains, but they're uneven and often difficult to measure. In reality, gains are smaller than the headlines suggest. The strongest gains are in bounded, repetitive, and information-intensive tasks — summarising research, preparing documents, supporting call centre staff. It's not simply the same work with fewer people. It's research with AI-assisted financial analysis, with humans in the loop, because you can have increased errors in forecasting in some circumstances. The productivity gains happen when an organisation can redesign workflows so that useful demonstrations of practice become reliable institutional performance over time.
Markets have rewarded anything labelled AI. Are we in a bubble or is this part of a long-term cycle?
I think both can be true. AI is a genuine technological revolution, but parts of the market are likely still overvalued. My assessment is that we are in the early stages of the technology cycle, but further along in the expectations cycle. That distinction matters because a technology can change an economy while investors are still overpaying for particular firms. The greatest risk is concentration — if a small number of companies and infrastructure providers account for much of the market's expected growth, disappointing earnings or slower adoption can create wider financial volatility.
Could AI leapfrog traditional financial infrastructure in emerging markets the way mobile did?
Yes — but it will not leapfrog infrastructure by eliminating its importance. It will leapfrog by making limited infrastructure more intelligent and more accessible. In these areas, what we need is AI to lower the cost of serving customers who have historically been expensive or difficult to reach — supporting local language financial services and alternative credit assessment. Mobile finance gave people access to an account. AI can give them access to personalised financial capability. The challenge is that the secret ingredient is trust, data, and public infrastructure working together.
As AI automates more tasks, which human skills are becoming more — not less — valuable?
The right question is what do we need to judge AI with. The human skills gaining value are often cited as critical thinking and analytical judgement, but you also need the energy of intellectual curiosity. The World Economic Forum reports that analytical thinking remains the most highly valued skill by employers. You also have to have the drive to keep learning. While AI makes production cheaper, it makes discernment and human trust much more valuable.
Nancy, that has been absolutely fascinating. Thank you very much.
My pleasure.
