AI gets talked about a lot as a buzzword, but my next guest has actually built it in. Görkem Köseoğlu is Group Chief Customer and AI Officer at Network International, leading customer product operations and AI strategy across the Middle East and Africa. Welcome to Wall Street to Mena. What brought Network International to Money 20/20?
This is where everything happens. As Network International we are the leading fintech in the region. We are processing hundreds of billions of transactions each year. We serve clients across 56 countries in the region. This is where we had to be — meeting clients and partners. It is a great event.
Can you tell us about Network International and your role as Chief Customer and AI Officer?
Network is a payment platform serving businesses, acquirers, and financial institutions — banks and fintechs. We have tens of thousands of business clients ranging from SMEs to large enterprises. We serve more than 250 banks processing their payments and payments-related activities. We position ourselves as a fintech and also provide value-added services beyond pure payments needs. We are active in 56 countries across the Middle East and Africa and are the leading processor in many of those markets.
You recently rolled out FICO Falcon for AI-powered fraud detection in Egypt. What does that catch that older systems could not?
Fraud-based AI systems go one step further beyond traditional rule-based systems. We use the analytics power of the systems to generate scenarios and fuzzy logic-based cases to catch new types of fraud. In traditional systems, you set certain rules about transaction types or patterns. With AI, we can identify outliers in a much deeper and more intelligent way. We are really helping our customers avoid fraud losses and respond much more quickly to potential cases. It is faster, deeper, more intelligent, and more accurate. One big area is false positives — we use AI to analyse transactions and come out with different thresholds or rules to reduce false positives so risk teams can focus on real fraud cases.
You also founded Prometheus AI. How far is the payment industry along in adopting agentic AI versus just talking about it?
Payments is essential — it is at the heart of many activities we do as consumers and for the economy. When it comes to agentic AI, we look at it across four areas. First, core payments — fraud management, improving authorisation rates. Second, AI-driven products for customers — agentic commerce, MCP protocol-based solutions. Third, tools for customers to interact with their own transaction data — a business can talk to an AI agent and ask how their transactions compare to last month or to competitors. That is very deep insight. Fourth, agentic payments is actually the last two seconds of the purchase journey — the whole consideration and decision-making happens before. Our clients want an end-to-end solution for agentic commerce, covering not just payments but how their products and services are discoverable in AI.
What does a more personalised, predictive payment experience actually look like for an everyday consumer or merchant?
Today a payment is a data point. Any payment has to be trustable, simple, and affordable. But beyond that, it is also generating a data point about your decision. For merchants and consumers, enriching that data and providing insights — what does my transaction pattern tell me, what does it tell about my business compared to competitors — is a great asset. Our solutions there we call data insights solutions or customised data solutions. For consumers, insights about spending habits, goals, and how to reach them provides additional empowerment through our banking partners.
What is the biggest risk of moving too fast when embedding AI across a payment business?
Trust is the single biggest component of AI going forward. Each AI capability we build goes through rigorous testing and validation — bias detection, explainability, transparency — both from a regulatory perspective and beyond it. For agentic commerce specifically, verifying that agents are taking decisions in the right way for the right users is an area we are working on with several partners. It is a fast-moving field. We are putting a lot of effort and governance into that piece.
Thank you so much for your insight today.
Thank you. Appreciate it.
