Societe Generale shares its strategy for integrating generative AI into the financial sector
Last month, Societe Generale sponsored the AI For Finance 2024 event in Paris, where it shared its vision and strategy for the integration of artificial generative intelligence in the financial sector. This annual event brought together leading European experts in AI and finance, highlighting strategic advances and use case demonstrations in the field.
Accelerating the bank’s transformation through Generative AI: A revolution in motion
Societe Generale's AI strategy aims to accelerate the transformation of the financial sector by harnessing the potential of generative AI to process unstructured data. According to Christophe Lattuada, COO of GBIS (Global Banking and Investor Solutions), this revolutionary approach opens up numerous possibilities and represents a significant strategic opportunity for the bank. However, it also emphasizes the importance of a balanced approach to simultaneously manage the risks, the costs and the support provided to staff impacted by this automation.
Optimized client experience and enhanced risk management: AI at the heart of Societe Generale's strategy
Noemie Ellezam, the Societe Generale group’s Head of AI, highlighted the current use of AI to enhance the client experience, with 300 use cases in production, 25% of which are directly aimed at improving the client experience.
Concrete examples: the use of chatbots and callbots to respond to clients’ most common queries queries, as well as the integration of AI to improve the relevance of offers and protect clients from fraud. SOBOT, the chatbot of SG Bank, handles almost 6,000 interactions a day and ELLIOT, the callbot of Boursobank, handles over 70% of client interactions from start to finish.
Finally, Christophe Tummers, Chief Data Officer of the Group, highlighted the challenges associated with managing data in a fast- changing IT landscape, emphasised the importance of the quality of data and the need to work together to develop AI frameworks and guides of best practices.
The experts' view
To learn more about Societe Generale's vision for generative AI, watch the fast video interviews with our experts Christophe Lattuada, Christophe Tummers and Noémie Ellezam.
Christophe Lattuada, COO of Global Banking and Investor Solutions
[Generative AI at Societe Generale]
[Societe Generale at AI For Finance]
[What is Societe Generale’s vision of generative AI?]
Our vision of the generative AI at Societe Generale is based on two strong convictions. The first one is that generative AI is a missing piece of the puzzle that will unlock a lot of value in terms of automation and digitalisation. And the reason for that is because basically it helps structure unstructured data. The second strong conviction is that it's not about mastering the latest technology, the latest Large Language Model. It's about using the whole suite of technologies to re-engineer our process to reinvent our activities. So, tt's about business transformation and business transformation at scale, instead of multiplying small experimentations.
[What are the main challenges and opportunities linked to its adoption?]
GenAI is triggering an acceleration of the transformation of our industry, and we shouldn't be afraid of it. First, because for years, we have been managing such transformation. GenAI is just another step in the digitalisation journey that started several decades ago. Yet, it's true that we need to manage the risks attached to it, and we need a robust control framework, a robust governance to make sure that we control the models, the automations that we will put in place. But also, we need to make sure that we accompany our staff in our organization. It's about change, it's about making sure that we build the capabilities, the skills that will enable us to make the most out of this technology and adapt our organisation and our businesses.
Christophe Tummers, Chief Data Officer of the group Societe Generale
[Data challenges in the AI era]
[Societe Generale at AI for Finance]
[What are your priorities to successfully enter the AI era?]
The main priority that we have is to enable AI in a very compliant way.
Obviously, as I'm very focused on data, it is all about making sure that we have good data quality, that we have known where the data comes from, and that we can protect the privacy of the data in there. These are essential preconditions to make sure that the data that we're using to train and use our models are sufficient and of the best possible outcome.
[How do you approach data governance within the Group?]
The way we look at data management is having a very, very strong data management framework. A data management framework basically describes the capabilities that we need to be successful from a data management perspective. We have data management framework 2.0 for Societe Generale, that we're currently implementing, which is an evolution of all the good work that has been done historically at the bank.
[How do you ensure data quality within the organization?]
One of the most important thing is that data and data quality is everybody's responsibility. Therefore, we have a very important books all about training and skills. We have rolled out a data-first program to make sure that we cover as broad a population in the bank as possible to transfer those data skills to everyone.
Noémie Ellezam, Societe Generale group’s Head of AI
[AI’s rôle in customer relations]
[Societe Generale at AI for Finance]
[How is AI changing customer relations within the bank?]
We already use AI for a couple of years actually, to enhance customer experience. So this can be directly in the interaction with the clients. For example, callbots or chatbots in the retail or in the capital market space. But GenAI is really bringing a new building block to what we can do to enhance customer experience with AI, notably in the way we are dealing with incoming customer queries, for instance.
[How can generative AI improve existing use cases?]
We can not only enhance the existing use cases that we have already, like chatbot, by better understanding, for example, the customer intention when they are calling us, but also to really display to them personalized customer journeys when we can have a full knowledge of the context of the clients and provide during the call, after the call, or to our front officers, advanced information about the customer to better serve their needs.
[What is your AI implementation strategy?]
What is the first step? The first thing is that we are selecting, six areas of priorities where we really want to focus our investments to deliver high impact use cases and very transformative process re-engineering based on AI and GenAI. And second thing is that we are working on an enterprise wide governance to really deploy an AI framework which will allow us to really manage all the risks, and there are multiples linked to the deployment of AI at scale.