As cyber leaders grapple with an evolving threat landscape, payments organisations face a unique challenge: identifying and stopping fraud before damage occurs. This panel 
explores how AI and machine learning can move beyond theoretical models to indicators of fraud behaviour in live payments systems.

Discussion points:

• How can AI models be operationalised to detect real-time fraud typologies in faster payments?
• What are the biggest barriers in integrating agentic AI into existing fraud engines and SIEM platforms?
• Balancing model explainability with performance: when should leadership accept ‘black box’ models and when not?
• Lessons from incident response: what successful AI-augmented detection looked like in practice.