Industry leaders have advocated for a careful approach to innovation using artificial intelligence, stressing the necessity of compliance, meticulous testing and building trust as fundamental for the effective introduction of AI and GenAI financial products and services into the market.
The comments came during this year’s Global Funding and Financing Summit, held at the European Convention Center Luxembourg, experts from the financial sector, including industry veterans and market leaders, offered a detailed and clear perspective on the future of artificial intelligence (AI) in the fast-changing financial sector. Speaking under Chatham House rules, which guarantees the anonymity of speakers and participants, they were particularly conscious of the growing oversight by market regulators and controllers.
The experts sought to clarify the reality beyond the excitement and hype surrounding AI. They explained that AI has been part of the financial sector for some time, used in various capacities. However, generative AI (GenAI), which allows the public to interact with applications based on large language models for text, speech and image tasks, has recently become sensational. Traditional AI and machine learning techniques are based on complex mathematical regression models, while GenAI offers logic-driven statistical outcomes which are more context-dependent. These new models, based on neural-network programming, are more like a predictive-modelling system, which the speakers agreed are akin to black boxes.
While the outcomes are not always fully explainable or understandable, GenAI, along with traditional AI, has become invaluable for tasks like fraud detection, document management, compliance and improving operational efficiency, to name a few, behind the scenes.
The speakers acknowledged that while regulatory approval can slow down the introduction of new technologies, it ensures that these innovations are socially responsible. They agreed that trust, safety and security are foundational for incorporating AI and GenAI into financial products and services, much like the trust required in banking.
To gain that trust, financial institutions and service providers must demonstrate the reliability, rationale and reproducibility of their AI-driven offerings. The discussion highlighted the importance of thorough and multi-step validation processes for these technologies, especially when the underlying LLMs and, therefore, the black boxes are getting bigger and increasingly complex.
The speakers agreed that AI and GenAI adoption by the financial world is inevitable but advocated for an overall very cautious approach to build trust. The speakers stressed the need for patience and investment to develop these technologies responsibly, considering their significant societal impact.
The summit attracted over 800 attendees, said a Deutsche Börse representative.
Under Chatham House rules, individuals can be quoted but they and their affiliation cannot be identified.