Artificial intelligence is experiencing rapid growth today with generative AI. But how does it concretely transform the banking system?
And why is it crucial to finance and support this innovation in France in the face of an eminently competitive international environment?
A technological breakthrough with systemic consequences
According to IDC(1), in 2024, banks have invested more than 150 billion euros in AI. Furthermore, according to the Boston Consulting Group(2), in 2025, 1 bank in 3 plans to invest more than 25 million dollars. Generative AI has only accelerated this trend over the past two years. But banks did not wait for 2025 or generative AI to integrate artificial intelligence or what looks like it. By recruiting data scientists, they began by capitalizing on big data and machine learning, then – to know their customers well, serve them better and make informed investment decisions – they gradually evolved towards increasingly sophisticated AI systems. New tools, capable of exploiting colossal volumes of unstructured data like never before: emails, calls, customer exchanges, social networks, economic press, social-economic data, extra-financial data, market sentiment, etc.
From immediate gains to structural changes
Banks use AI for 4 main types of uses. First of all, improving the customer experience. The objective is to optimize marketing and sales via virtual assistants (chatbots, voicebots) and recommendation and personalization systems, in particular thanks to generative AI. Second use, the optimization of internal processes to free advisors from low value-added tasks: automation of manual and repetitive tasks, processing of unstructured data and empowerment of clients for day-to-day operations. Third key use, risk management such as fraud detection, anti-money laundering and anti-terrorism financing (AML-FT) as well as its corollary: customer knowledge (KYC), management of alerts and financial, operational (including cyber risks) and compliance risks. Finally, fourth axis, market activities where AI is used for predictive analysis of market trends, assistance with portfolio construction and/or asset valuation.
A binding framework that shapes innovation
The European AI Act classifies AI systems according to their level of risk. However, the banking sector, through its customer profiling/credit scoring activity, finds itself placed in the high-risk category. To guarantee ethical and trustworthy AI, additional requirements must be respected (data quality, explainability of decisions, human supervision, complete documentation, etc.). Banks must therefore provide the mechanisms and traceability necessary for transparency of decisions taken by AI, without forgetting to respect the various regulations – GDPR, DORA or NIS2. Heavy constraints which have a cost and weigh on their competitiveness, particularly in Europe and at the international level where competitors operate within a much more flexible framework.
Investing in AI to reinvent banking
In France, in terms of financial support for innovation and research, we have two key tools. First of all, the Research Tax Credit (CIR), with an annual budget of 7 billion euros. The other system is “France 2030”(3), with a budget of 54 billion euros over 5 years, 40 billion of which have already been invested since the end of 2021. Each of these programs devotes a significant part to digital and AI. The use of the CIR by banks, in the same way as by large industrial or pharmaceutical groups, regularly arouses criticism, on the grounds that, these players already having significant financial resources, the system should primarily target SMEs.
Make social responsibility a priority financial focus
According to the latest data from the Ministry of Higher Education, Research and Innovation, the banking sector – all players in the financial sector included, institutions, but also insurance companies, asset managers, etc. – accounts for less than 2% of the annual CIR debt. No longer supporting innovation in this sector would therefore send the wrong signal to our financial industry, currently engaged in costly and risky AI projects (explainable AI, regulatory compliance, HPC and secure cloud infrastructures, cybersecurity, fight against fraud and financial crime). Undoubtedly the price to pay for an ethical and trusted AI unique in the world, in a sector where trust is the most precious asset.
A lever of positive impact on the entire ecosystem
In addition, banks outsource a large part of their R&D to laboratories, ESNs, fintechs, generating a positive knock-on effect on ecosystem innovation. A major advantage is that many French banks now even have internal incubators or accelerators to identify, test and integrate innovations. Some even go so far as to integrate promising start-ups into their range of services. Finally, the competitiveness of banks is also a national and European issue, even of financial sovereignty, knowing that in other geographical areas, the financial support of the State is significantly stronger.
Not investing in banking AI means risking losing control of our financial sector in the face of international competitors who do not face the same constraints. France has several banks in the world’s top 10, including 5 in the top 50 in terms of AI adoption.(4). Preserving this position requires strong public support and collective mobilization.
(1) International Data Corporation (IDC) Worldwide Artificial Intelligence Spending Guide
(2) BCG, For Banks, the AI Reckoning Is Here, Financial Institutions Ai In Fi Report 2025
(3) France 2030 Plan https://www.economie.gouv.fr/france-2030
