Yusuf Yassin: How is Banque de Luxembourg adopting AI in its processes? What impact will it have on day-to-day experiences for employees internally and clients externally?
Stephane Bouvy: Since 2018, we have had a department focused on data utilisation. This enables us to work across the entire value chain of data exploitation, from descriptive to cognitive. AI thus represents an additional way of exploiting data.
We don’t do AI for the sake of doing AI. Our positioning is to use AI in a reasoned way, where it makes the most sense in our private banking activities, and always within an ethical framework that protects the integrity of our data. AI is not intended to replace human intelligence; on the contrary, it complements and reinforces it.
Currently, our team of data scientists is working on several use cases, enabling us to constantly propose new functionalities and evolve our approach over time.
We also support two PhD students, one at the University of Luxembourg and the other at Maastricht University. At Banque de Luxembourg, our vision is to prioritise constant improvement in our processes and to contribute to this dynamic, evolving ecosystem in Luxembourg and particularly in the financial system.
Do you have any use cases you could share with us? What is the goal, and how will it work?
We have several use cases in the pipeline at the moment. These are essentially to support our business processes and anti-money laundering compliance. In general, these use cases are part of an approach designed to facilitate the work of our employees.
While remaining the customer’s privileged contact, and building on the relationship of trust built up over time, the private banker can be supported by AI, depending on the case and the customer’s needs. In particular, this reasoned and supervised use enables them to be relieved of repetitive and time-consuming tasks, so that they can focus on the intrinsic value of their mission.
We also have a fraud detection model for transfers that has repeatedly demonstrated the relevance of these technologies. This enables us to protect our customers and facilitate the work of our employees. This approach is therefore virtuous on two levels: for our customers and our employees.
While some data ‘prophets’ are already predicting that AI could replace board members by 2030, we believe that human intelligence will continue to play an inescapable role in the decision-making process and in general.
What steps are Banque de Luxembourg taking to conform with the European Union AI act in its current form? What are the biggest challenges and opportunities?
The aim of this regulation is to guarantee better conditions for the development and use of this technology, and to provide greater transparency and protection. This solid framework aims to strike a balance between promoting innovation and ensuring the responsible development of AI. To achieve this, the text relies on a risk-based approach, categorising AI systems according to different levels of risk.
This desire on the part of the regulator to provide a framework for practices and thus avoid aberrations is a good thing. It should be remembered that these regulations apply to all areas of activity, not just the financial sector.
The potential of this technology remains impressive, and it seems important to us to guide the various players in the direction of consumer protection.
With this in mind, we have undertaken a review of our existing models in order to categorise them according to the regulations and prepare for their implementation. Our reasoned and targeted approach to the adoption of these technologies puts us in a good position to welcome this new regulation, even if further efforts will be required.
Looking to the future, how will AI change the customer’s banking experience in 10-15 years? How will it change compared to today?
It’s not easy to project and estimate the potential and impact of AI in the financial industry. But AI is an opportunity and a tremendous help in business development.
While some data ‘prophets’ are already predicting that AI could replace board members by 2030, we believe that human intelligence will continue to play an inescapable role in the decision-making process and in general. AI will help us, at different levels and in defined use cases, but humans will and must remain the ultimate decision-makers. It's important that there is no AI without human intelligence. And this is all the more true in the field of private banking, where the relationship with the customer is precious, and where trust is a pivotal aspect of the relationship.
But it’s clear that these technologies will profoundly change the way we work in the sector. We have an additional tool at our disposal, and it's up to us to use it responsibly and ethically.