AI: We can walk the walk, but can we talk the talk?

An interview with Thierry Derungs, Chief Digital Officer at BNP Paribas

Following the AI Summit day at FinovateEurope, we sat down with Thierry Derungs, Chief Digital Officer at BNP Paribas Wealth Management to discuss AI within the enterprise and the process to changing a company mindset in order for AI to have a true impact.

What first sparked your interest in how AI can improve financial services, and what has changed since then?
AI is nothing new: for a long time, companies have been trying to improve their efficiency, speed and accuracy through automation. What has drastically changed is our affordable computing power and the accessible data volume - the two AI fundamentals.
My very first interest in AI was to move out of the truncated view a “classic” algorithmic or analytic approach could provide. Indeed, going in more depth to embrace more data and increase accuracy goes along with the exponential complexity we’ve seen in recent years. Today, the explosion of AI-accessible technologies moves our crossing of oceans from caravel boats to aircraft: it’s affordable, available to many, faster and more manageable.What is exciting you most at the moment when it comes to applications of AI? These 2 simple letters hide a whole wide new world, and today we are still only at the edges of what’s possible. I am fascinated by this rapidly-evolving world… Nevertheless, I must say that my fascination goes along with some concerns. I believe that digital has never been so human related and I always worry about our change capacity to embrace these new opportunities.

Change management, in terms of soft skills, must never be neglected nor underestimated - especially when the pace of new technologies goes much faster than our usual human adaptation capacity.

Neil Armstrong walked on the moon thanks to a thousandth of the computing capacity of the first iPhones. Today, many people would be concerned ‘only’ to have that same iPhone’s GPS to find a restaurant around the corner… People’s trust in technologies mainly comes from a sense of ease. I consider it my duty to guide our employees and open their understanding, raise their awareness, to allow the best usage.The title of your presentation at FinovateEurope was ‘AI: We can all talk the talk, but can we walk the walk…?’. Can you tell us a little more about what inspired the title? Listening to many consultancies, you’ll often hear that if you just put some AI box in place, magic will happen - and that if you don’t have AI your business will be blown away. As a result, I see many enterprises jumping on the topic with huge expectations, while dedicating too few resources due to the plug & go illusion.
My presentation set out to share our experience as, indeed, there are many opportunities, but they require a step-by-step approach: a good learning curve combined with fast enough movement. From neural analytics to behavioural fraud detection, I hopefully showed that working in agile ways, learning through an MVP inside our Factories (our internal incubators and accelerators) and collaborating closely with startups, are good ways to move forward with AI.AI presents us with new approaches to detecting and preventing fraud before it happens – how is behavioural analytics being used in your team? Security and fraud have always been a must-have for BNP Paribas. They are our duties towards our clients and are a strong focus at every level - from our CEO to our IT department.

Any five-year prediction is very risky considering the current fast pace of AI evolution.

The combination of behavioural analytics together with machine learning allows us to build new individual detection patterns while having the capacity to adapt individually. You and me, even if we look alike, are very different: so, too are our habits. AI gives us the capacity to look differently to each individual and constantly adapt, and strengthen our fraud prevention and detection abilities in terms of accuracy and speed.How do you put in place the change in teams’ mindsets that’s needed for AI to have an impact? Change management, in terms of soft skills, must never be neglected nor underestimated - especially when the pace of new technologies goes much faster than our usual human adaptation capacity.
Our step-by-step approach not only allows us to learn but also to demonstrate a tangible added value. For example, to convince on neural analytics, we did a competition between the classic and the new solution. Both had to build the same model with the same means, and we then tested both models in production: neural analytics showed even better results than expected, so the added-value was proven and people were convinced.
I love in particular two quotes from Einstein: “It’s harder to crack prejudice than an atom”, and “God gave me the stubbornness of a mule and a fairly keen scent”!
Awareness, change management and clear explanations based on business use cases are the best ways to move from a “No” to a “Mmmh” then a “Why not” and finally the expected “Let’s try”!How do you see AI use in BNP Paribas evolving, and what initiatives would you like to see in place in 5 years’ time? Any five-year prediction is very risky considering the current fast pace of AI evolution. Of course, I could raise the catch-all, and vacuous, statement that “AI will be key”, but I think we all know that.
What is trickier is to predict for what and how far AI will be useful to support our promises to our clients and fulfil their expectations.
We have a business vision to achieve, and we constantly challenge ourselves for each initiative though our clients’ involvement from the very first concept to use in production. Based on client feedbacks we adapt when requested what we do, what we want to do and our longer-term vision. AI, together with other new technologies, will be surely a key asset to support our engagements to our clients.

Check out the FinovateEurope highlghts >>