Innovation and Digital
AI is no longer just an option in finance

AI is no longer just an option in finance

Data, algorithms, Machine Learning, bots, a tech vocabulary adopted by all segments of the banking industry around the world, all for the customer’s benefit.  

The massive digitization of the economy, accelerated by the health crisis, has provided the various segments of the banking industry with multiple opportunities to generate value for their customers: increased digitization of products and services, personalization, higher service quality and development of new business models. To better serve their customers, all of Societe Generale’s businesses are constantly working to ramp up their digital transformation, in particular by adopting data processing and artificial intelligence (AI) technologies.

Data is a key asset for the bank, AI is just a tool

The digitalization of processes and interactions with customers has significantly expanded the pool of data at the bank’s disposal. The way data is used is a key component of the digital transformation in the financial industry. 
The analytical and predictive capabilities of Artificial Intelligence contribute to extracting the full value of these data, in compliance with personal data regulations. AI is thus a tool that boosts the digital value proposition offered by Societe Generale businesses to their customers. 

Elliot (Boursorama): 4 million conversations per year

AI provides detailed knowledge of customers, which is useful in anticipating their expectations, giving advice, and making personalised and appropriate business proposals. In Romania, for example, BRD, the country’s leading private banking network and a subsidiary of the Societe Generale Group, uses AI to create marketing campaigns targeted to customers with the greatest appetite for consumer loans. The result is a threefold increase in loan acquisition rates.  
AI is also used to handle simple customer requests faster and more efficiently through automatic natural language processing1. This feature is capable of understanding requests regardless of the means of communication employed (written or voice) and developing high-quality conversational interfaces. For example, Boursorama processes more than 4 million conversations per year thanks to AI with the Elliot conversational engine. 

AI: the age of reason

Today, AI is fully incorporated in all Group businesses, which have now achieved a certain maturity in this area. Societe Generale has identified 130 AI technology use cases in the fields of customer service or operational efficiency. The Group recently won the 2021 FSTech2 Award for the Best Use of Artificial Intelligence (AI) in Financial Services.
The key challenge now is to industrialise these solutions, i.e. to expand and replicate existing use cases throughout the Group, where relevant to customer specifications and regulations governing the Group’s activities.
One good example of this is MOSAIC (More Security with Artificial Intelligence), our solution for detecting fraud in the use of payment instruments, first rolled out in France for individual and corporate customers. MOSAIC is based machine learning3 and behavioural analysis algorithms to identify abnormal or suspicious events and trigger warnings where appropriate. With this solution, fraud can be detected on an instant payment in less than half a second. MOSAIC has been broadly deployed in France, and is now also offered to the Group's subsidiaries in Africa and Europe. 

No AI without human

Data and AI cannot be used to accelerate the digitization of our services without people. People are the ones who design, produce, implement and use these AI algorithms or solutions. Employee expertise is both central to the value chain and keeps it moving forward. 
That is why the entire company needs to be on board in order to accelerate this transformation and guarantee excellence in execution, which calls for a paradigm shift involving new working methods, adoption of a data-driven culture and a real focus on the complementarity of skills and talents.

[1] Automatic natural language processing mimics human understanding of words and phrases and now allows machine learning models to process large amounts of information before giving precise answers to questions asked.
[2] FSTech is a leading web-based media on technology in financial services in the UK and EMEA (Europe, Middle East, Africa).
[3] Machine learning is a field of study of inartificial intelligence, which employs mathematical and statistical approaches to give computers the ability to "learn" from data.