This article has been reviewed according to Science X's editorial process and policies. Editors have highlighted the following attributes while ensuring the content's credibility:
The AI-driven evolution of banks
AI-driven transformation continues to accelerate. In this phase, banks are faced with new challenges and opportunities, which force them to pay attention to a plurality of aspects, all transversally crossed by a major cultural transformation.
The main aspects affected by this transformation are, on the one hand, corporate strategy, which concerns the degree of pervasiveness of AI within banking activities, the dedicated budget, the benefits encountered and the challenges that remain open, as well as of course their progress over time. On the other hand, however, there are the governance aspects, which embrace organizational issues, processes, management of skills and issues relating to risk management and ethical implications.
With reference to the strategic aspects, it should be noted that AI today represents an increasingly important investment item for Italian banks. 76% of respondents to the ABILab survey (2023) state that they have a dedicated budget and in 65% of cases this budget is more than €500,000 and 59% of those interviewed expect that budgets will further expand in 2023.
There is a strong tendency to interpret AI as a strategic driver of business transformation. In fact, 88% of respondents stated that they have defined/are defining an ad hoc strategy, which will then have to be put in synergy with the company's Data Strategy. This theme also implies the strengthening of partnerships and ecosystems useful for promoting innovation.
In fact, Italian banks are already using mixed sourcing models for the development of AI solutions, using make and buy levers accordingly and interfacing with a variety of players (ICT partners, startups, etc.)
In defining an AI-driven transformation path, it is also important to frame the benefits that banks intend to seek and obtain. In this regard, it is interesting to note that answers to the survey do not exclusively refer to higher economic returns from a reduction in costs (53% of the sample), but also to improvements in decision-making processes (53%) as well as strengthening the relationship with the end client (41%) and, last but not least, improving existing products/services.
In fact, the majority of use cases already present on the market concern support to (assisted and self-service) channels and support to control functions.
On the other hand, as regards governance issues, most of the responding banks (58% of them) say they are working on implementing a framework for the governance of AI and these initiatives are often placed in continuity with the efforts that have been devoted in recent years to the definition of a company model for Data Governance, now widely operational in banks.
The main pillars on which the banks intend to build this governance system are those of transparency, explainability and traceability. They also believe it is useful to have a monitoring system that makes it possible to evaluate the results of AI from the point of view of business effectiveness, technical efficiency, governance and ethics. This kind of monitoring that is already in place in about 40% of the sample.
A further important factor concerns skills, which are necessarily heterogeneous (as they range from risk management to cybersecurity and privacy) as well as the need to hire new profiles in the AI field with reference to Data Scientists, Data Engineers and Machine Learning Engineers.
In the area of governance, another relevant issue concerns the verification of ethical principles, which refer to the concepts of fairness (ensuring that AI is fair and impartial), transparency (verifying how data is used and how systems make decisions) in addition to privacy aspects (that is, aimed at ensuring that customer data are not used beyond the intended and declared use).
Last but not least, banks also intend to invest in accountability, highlighting the rules, policies and models that make it possible to determine responsibility for the decisions taken by the AI system. Naturally, in all of this, a significant role will also be played by regulatory safeguards regarding the compliance of AI systems.
As a result, new challenges arise for banks. Bearing in mind that AI will not only concern technological skills, but will introduce changes throughout the organization; on the governance front, organizational processes and systems will need to be put in place to favor its conscious and responsible adoption. Furthermore, the speed of adoption of AI will be a key factor, as well as the understanding of its impacts on the ethical, social and sustainability spheres.