Key learnings from the panel discussion with Johan Bryssinck, Program Lead for AI/ML at Swift, Chalapathy Neti, Head of AI/ML Platform at Swift, and Tobias Hann, CEO of Mostly AI, at the AI for Financial Services Summit by Artefact - June 12, 2024.
About Johan Bryssinck: At Swift he leads the adoption of artificial intelligence to enhance products and services. With over 20 years of experience, he focuses on corporate strategy, innovation, and technology, driving partnerships to solve industry challenges using AI.
About Tobias Hann: He is specialized in synthetic data and AI-driven solutions. He holds a PhD from the Vienna University of Economics and Business and an MBA from the Haas School of Business, UC Berkeley, and has extensive experience in software, data, and machine learning.
About Chalapathy Neti: At Swift he develops enterprise-scale AI platforms. He has extensive experience in AI, cloud solutions, and has held senior roles at IBM, including VP of IBM Watson Education and Director of Healthcare Transformation.
Introduction
The discussion focuses on the significant challenges within the financial services industry, particularly concerning financial crime, fraud, and anti-money laundering. These issues are estimated to cost around $480 billion globally. Tackling these problems at scale requires collaborative innovation, primarily because data is dispersed across multiple silos.
Role of synthetic data
Synthetic data emerges as a crucial enabler for this collaborative innovation. Swift, a dominant payment rails company, facilitates a significant volume of global GDP transactions and collaborates with numerous institutions across the globe. To address financial crime effectively, Swift collaborates with partners to utilize synthetic data for innovation responsibly.
Approach to financial crime
Johan elaborates on the persistent issue of fraud, emphasizing the need for collaboration to combat this problem. Despite technological advancements, fraud continues to rise, partly due to the fragmentation of payment methods and the existence of data silos. Swift, in collaboration with the Future of Financial Intelligence Service, has observed significant improvements in fraud detection and prevention through data sharing initiatives.
Anomaly detection and AI
Swift’s ambition is to build an advanced anomaly detection model for real-time transaction monitoring. Collaboration with banks and leveraging AI in payment control services have already shown promising results, such as a 40% reduction in false positive rates. The next step involves integrating anomaly detection in payment prevalidation services to enhance transaction security before initiation.
Confidential computing
Confidential computing is highlighted as a key technology for secure data collaboration, allowing for data protection during all stages of processing. Swift aims to scale this technology globally, working with hyperscalers to reach its extensive customer base.
Importance of synthetic data
Toby from Mostly AI discusses the significance of synthetic data in enabling secure and effective data collaboration. Synthetic data, which is fully anonymous, helps in various use cases such as software development, research, and training AI models. It also addresses data privacy concerns and aids in creating more robust and unbiased AI models.
Challenges and future directions
The discussion also touches on the impact of AI on employment and the emergence of new payment companies as potential competitors. Swift continues to innovate by integrating new technologies and expanding its payment rails to adapt to changing financial landscapes.
Conclusion
The conversation underscores the importance of collaborative innovation, AI, and synthetic data in addressing global financial crime. Swift’s ongoing efforts in anomaly detection, secure data collaboration, and synthetic data utilization are pivotal in enhancing the integrity and efficiency of the financial ecosystem.
