AI is changing what it means to be a bank
We are probably only at the beginning of a revolution driven by generative artificial intelligence. In which sectors can we expect the most significant developments from new AI tools in the future?
The impact of AI already affects and will increasingly influence all areas of banking activity. Even today, the use of AI tools is widespread and constantly expanding, both in customer interactions and in areas related to the bank’s internal operations.
How does AI fit into Banco BPM’s digital transformation strategy? What AI solutions is the bank developing along this path?
Banco BPM is making strategic investments in AI, integrating it into commercial, operational, and risk management processes, with the aim of improving service quality, increasing productivity, and supporting employees’ professional growth. For example, AI is used for customer profiling, identifying the most suitable products to meet their needs, and accelerating and improving creditworthiness assessment and loan disbursement processes for businesses and households. AI is also used in commercial activities. In this context, the bank has developed Paolo, a virtual assistant (both chatbot and voicebot) that uses machine learning technologies to provide fast, high-quality support 24/7 to customers who need information or want to carry out simple operations that do not require advisor support. Through Paolo, the bank handles more than 75% of customer service requests made via mobile channels. Paolo has also become a tool for proactively delivering commercial messages. Integrated with the marketing automation platform, it can identify whether a customer belongs to the target of specific campaigns and, if so, suggest relevant products or services and propose further engagement through the digital branch, including scheduling calls with expert advisors who can also complete remote sales. AI is also used in middle- and back-office activities, automating repetitive processes, reducing errors and risks, and increasing productivity by at least 20%. It supports staff in branches as well, particularly in anti-money laundering processes, helping to identify false positives in suspicious transaction reports, reducing unnecessary alerts, and improving overall system effectiveness. Among the areas where the most significant innovations are emerging, in addition to those already mentioned, regulatory analysis, interpretation, and application are particularly noteworthy. In a highly regulated sector like banking, subject to laws, regulations, and multiple authorities, generative AI is enabling important improvements in efficiency, speed, and cost reduction. To support employees, the bank has developed a conversational chatbot that allows them to access regulatory information more easily and quickly. Staff can interact with the system using natural language to request information or assistance on specific operations. For example: “I need to open a current account for a customer. What procedure should I follow?” or “I need to process a mortgage application for a non-resident. Which documents are required?” The application searches its trained knowledge base (internal policies and external regulations) and provides concise answers in natural language, also indicating the sources used so employees can verify accuracy and relevance. The chatbot can also respond to questions related to legal opinions and HR processes. AI also plays an important role in relationships with businesses, particularly in credit evaluation and lending processes.
How do people interact with these technologies?
People remain and will always remain central, especially in commercial and credit processes. Typically, AI applications support staff in carrying out tasks and making decisions, but they do not replace them. Technology may suggest an answer, but the final decision is made by those managing the customer relationship. This happens consciously, as employees understand both the data used and the logic behind the algorithms. In credit approval processes, for example, if an application is rejected, the system provides the reasons so that the advisor can return to the customer and adjust parameters (such as extending repayment periods, adding guarantees, or slightly modifying the loan amount) to make the request acceptable. Even in these cases, human intervention is always required for the final decision.
What impact do you expect on employment?
While some roles may decline, demand for new skills and professional profiles will grow, creating new job opportunities. From a qualitative perspective, the mix of roles will change, with an increase in highly specialized positions. As mentioned, Banco BPM uses AI to support staff, whose contribution remains essential: human intervention is always required for final decisions. This approach also helps reduce the risk of discrimination in AI usage. To support employees in adopting these technologies, the bank is implementing targeted training programs. It has also established an internal center of excellence with highly specialized resources to drive AI implementation.
The level of investment required to fully leverage AI seems significant. What can we expect?
Adopting AI in banking requires substantial investments, both in technology and skills. In its 2024–2027 business plan, Banco BPM plans to accelerate its digital transformation with €800 million in IT investments, focusing on initiatives aimed at fully digitizing customer offerings, improving service quality, and increasing operational efficiency through process simplification and automation. Within this framework, the bank is implementing a progressive AI adoption program to transform its way of doing banking, leveraging advanced analytics, machine learning, and generative AI to support revenue growth, cost control, and credit and risk management. These investments represent a significant commitment, and the size of a financial institution plays a key role in sustaining them. An alternative approach is to build partnerships with specialized companies. In fact, it is not always convenient or efficient—even for large organizations—to develop all technological solutions internally, especially in rapidly evolving fields like AI. There is significant potential for collaboration with fintech companies, which can be successfully explored by institutions of all sizes.