Artificial intelligence: from concept to reality in investment funds

AI at a glance

Artificial intelligence began in the 1950s when the first artificial neural network was created by two students from Harvard University. These networks have become an important part in the categorisation of AI that were subsequently developed. Two major essential concepts then emerged: general AI and specialised AI. The first is intended to be a multi-task life assistant while the second is used to perform tasks in a specific field with an increased degree of accuracy.

In terms of operation, we can distinguish between two major families of technologies: Machine Learning and Deep Learning. Machine Learning is a data analysis mechanism through algorithms to identify models and make predictions. Deep Learning is a branch of Machine Learning that focuses on the use of artificial neural networks. Machine Learning is therefore an advanced form of data analysis whereas Deep Learning aims to copy the functioning of the human brain.

The application areas are multiple and use both Machine and Deep Learning. Examples include Natural Language Processing (NLP), robotics, expert systems, reinforcement learning or computer vision. Respective applications include Chat GPT and Alexa, Boston Dynamics or robot vacuums, the IBM Watson medical consulting system, video game bots and facial and voice recognition.

Lastly, a final major concept is that of generative AI. This term refers to an AI's ability to generate data based on content created by humans. Its operation is based on a generation and filtering system that enables it to improve continuously when exposed to a comparative database.

Why is it such a success?

Thanks to this wide range of uses, AI is finding a place in all sectors of activity, and in particular in the field of investment funds, where margins are often reduced.

For companies, it allows the automation of many repetitive intellectual tasks with low added value and thus frees up time for projects with higher added value. The direct consequence is a reduction in costs and an increase in productivity, as generative AI produces large amounts of content in a very limited time.

In addition, this technological revolution appears to be an opportunity for social progress. The increase in productivity, lower costs, less repetitiveness at work and increasingly autonomous working tools could lead to an increase in living standards and a decrease in the hardship of work.

The recent launch of Chat GPT by OpenAI has had a major impact on the public perception of artificial intelligence. The fact that the application is free has led to massive adoption by users, who have been able to discover the state of the art of the technology. This has been followed by an explosion of usage ideas on social networks and many new business models are starting to emerge.

Our progress at Societe Generale Securities Services (SGSS)

To date, the Societe Generale group's portfolio has more than 330 Data and AI use cases (UC) in production, 170 of which are based on AI, in order to execute the strategy with an expected value creation of €500 million by 2025.

At SGSS, we take advantage of the full potential of these new technologies and the strength of our Group to offer the best possible services to our clients. The search for efficiency and automation of our processes has become a real leitmotiv, just like client experience or business efficiency.

For example, our new intelligent filtering tool enables us to identify unusual cash flows and respond more effectively to our anti-fraud and regulatory control obligations.

Based on our in-house digital capabilities, our automatic prospectus reading tool enables us to speed up the client onboarding process by capturing key information from documents and integrating it directly into our database.

Last example, which allows us to further improve the speed of work for our clients, a tool that summarises the fact sheets on the performance of our client funds to enable automatic analysis and processing of information that further reduces the cost of our services. 

In general, the AI component is an integral part of SGSS's new production tools. We are now in the "AI by design" era.

These advances and the collaborative work of Man – Machine strengthen our data management with the aim of positioning ourselves on the key issues of the future. AI is at least as much about culture as it is about technology. At Societe Generale, we very quickly understood the need to train all employees, in a variety of ways, from awareness-raising (a few hours) to further training (a few days), as well as retraining (a few searches and a change of job at the end). Good control of data and technologies guarantee the innovative services of tomorrow for our clients and new generations.

Laurent Marochini, Head of Innovation, Societe Generale Securities Services Luxembourg