en de

Online Magazine

How can banks become truly AI-driven?

In today's constantly evolving business environment, it is essential for banks to leverage data analytics, automation, and AI to keep abreast of industry trends. However, merely implementing individual use cases is not sufficient for maximizing the benefits of these advanced technologies. Instead, banks must strive to become AI-driven at every level of the organization.

By Nita Ngoy-Tha

The year 2023 presents a multitude of challenges for financial institutions, particularly banks, as they grapple with the ongoing impacts of the COVID-19 pandemic, the war in Ukraine, and soaring inflation. Additionally, they are faced with the need to keep up with rapidly evolving industry trends, including increasing customer centricity, a more restrictive regulatory environment, and rising pressure on operating expenses.

How, you might ask, can banks fulfil these expectations in this ever-more challenging environment? Let us look at what the current industry trends are about, and how data, automation and artificial intelligence (AI) can support banks in living up to them.

How data analytics, automation and AI help banks can keep up with industry trends

Trend 1: increasing customer centricity
Customers increasingly demand banks to adapt their services to their needs and preferences. On the one hand, they want to use financial services and get the necessary support whenever and wherever they feel like it. On the other hand, customers are holding banks to high environmental, social and governance (ESG) standards. With data analytics, automation and AI, banks can increase their customer centricity in the following 3 ways:

  • Personalized service: Banks can leverage technologies like the metaverse and virtual reality to connect to their customers in new ways and thus make the entire banking experience more personalized. Moreover, by means of the metaverse, banks can meet the growing demand for digitally native currencies.
  • Increased service speed: By using AI and machine learning (ML) for upselling or cross-selling, banks are faster in developing innovative products and services and bringing them to the market.
  • Social responsibility: AI can help banks promote sustainability and other social responsibility themes by spotting investment opportunities that are green and at the same time bring maximum profit returns.

Trend 2: an ever-more restrictive regulatory environment

Together with healthcare, the finance industry faces the highest regulation costs. Banks are required to know their customers, uphold customer privacy, monitor wire transfers, prevent money laundering, etc. On top of this come the potential costs of regulatory fines and penalties. Data analytics, automation and AI can help reduce these costs in the following 3 areas:

  • Smart risk management: Data analytics and AI can help banks monitor events that might impact their daily operations, such as economic and political developments, compliance to ESG requirements, or decisions by competitors. AI can predict outcomes in advance and help banks protect themselves against short-term losses while developing more accurate risk models.
  • Fraud and cybersecurity threats prevention: To save regulatory fines due to customer data loss, banks can use ML algorithms to detect suspicious activity on their own servers or on customers’ accounts.
  • Identity management automation: Remote identity verification applications use facial biometrics, extract identity documents, and run authenticity checks to identify individuals. With them, banks can automate their identity management processes which makes them more cost-effective.

Trend 3: rising pressure on operating expenses

While banks are already under pressure to lower their operating expenses, the current value loss of money due to the inflation drives up these costs even more. Here are 2 concrete ways in which robotic process automation (RPA) and AI can help banks cut operating expenses:

  • Intelligent operations: A big part of operating costs are caused by time-intensive and error-prone work such as entering customer data from contracts, forms, and other sources. Automated workflows make these back-office tasks leaner and thus more cost-efficient. E.g., bots can automatically update client contact data across various systems (automation technologies include UiPath, Blueprism or Automation Anywhere).
  • From legacy to agility: Banks can transform their core systems into intelligent, connected, flexible systems using the agility of the cloud. Moreover, the cloud can act as single point of truth to gather internal or external partner data and derive insight generation.

As we can see, data analytics, automation and AI have great potential to support banks in adapting to new industry trends. However, merely implementing individual AI solutions is not enough and can even lead to undesirable outcomes.

For example, it is important to decide in which cases an automated chatbot makes sense and in which cases it does more harm than good. Similarly, banks need to determine how much of the risk management they want to automate and where they still need to include experts. In the context of operations, using cloud solutions can have many benefits, but this too needs to be approached strategically and in accordance with a bank’s business goals.

Thus, to get the most out of data, automation and AI technologies in the long term, banks have to integrate them holistically in their organization – from their strategy to their operating model, organizational readiness, data foundation and technology. Let us check what that looks like in concrete terms.


How will rising interest rates affect banks' product innovation?

Why should banks start treating data as a product?

And to what extent will individual branches play a greater role again in the future?

Find out all this and more here: in the Banking Trend Report for 2023.

How banks can become truly AI-driven

Banks that are interested in becoming AI-driven need to adopt an AI-first mindset and build a holistic set of capabilities. To do so, banks should consider the pillars shown in Figure 1:

Figure 1: Pillars in which to integrate AI to become an AI-driven bank.

AI Strategy: Banks need a clear data and AI strategy that is aligned with the overall bank and IT strategy and guides their transformation into an AI-driven business. As a first step, a bank has to answer questions like: How can we assess our data and AI maturity? Where do we stand in comparison to our competitors? On this basis, it can then develop the related strategy and roadmap to reach the target state.

Business Value & Analytics Use Case Portfolio: In the context of this pillar, granular AI-driven decisions are required. Individual data analytics, automation and AI applications – such as a chatbot or an automated fraud detection dashboard – contribute to the bank’s business goals and generate concrete value.

Operating Model: Besides a data and AI strategy and specific AI implementations, banks need to integrate AI in their operating model. This includes deciding how centralized the AI capabilities should be. Also, they have to include new ways of working, such as Agile and DevOps, to make the delivery of a business relevant use case agile and scalable.

Organizational Readiness: Employees need to be able to work with AI-driven solutions. Therefore, banks have to further data literacy in all functions of the organization, e.g., by providing data literacy trainings. They can even make the fulfillment of these trainings part of their employees’ performance criteria. Also, when recruiting, banks can make basic data literacy a criterion – regardless of the candidate’s role.

Data Foundation: Well-structured data management (the ability to access, ingest, and transform data securely across the organization) is key for banks to get the most out of their data. Data management includes, e.g., data governance with clear data ownership, as well as the maintenance of data quality and data security. Once a solid data foundation is in place, the implementation of data-driven and AI solutions is much easier.

Technology: Banks’ products and interfaces need to be able to seamlessly integrate with ecosystem partners via API. This allows customers to discover and consume propositions beyond the bank’s core platforms. The easiest way to achieve this is to move to the cloud, as it provides a single point of truth to gather data as basis for data and AI use cases.

Considering data analytics, automation and AI across all pillars in the organization can enable banks to go beyond the initial AI experimentation phase, scale and become thoroughly AI-driven. With this foundation, they have the best chance of keeping up with the initially mentioned industry trends.

But what benefits does this even have in a concrete use case? In the following example, one of our teams helped a global banking institution to establish a data platform and thus find new revenue streams.

Use Case: Fighting pressure on operating expenses with a data platform revenue stream

Like many banks nowadays, one of our clients, a global banking institution, experienced increasing pressure on its operating expenses and thus had to create new revenue streams. To do that, they knew they would have to transform their core banking business model. Meaning: make it more data driven.

To support our client on this mission, we explored and tested several data-driven business models that would potentially help the bank monetize their services more efficiently. They were all so-called “open banking business models”. They all entailed the bank’s sharing their banking data through APIs with two or more unaffiliated parties to deliver enhanced capabilities to the marketplace. The business models we tested were the following:

  1. Direct monetization of API-calls
  2. Monthly flat fee for service usage
  3. Additional fee for developer portal access
  4. Indirect monetization through direct investment in partners

The fourth open banking business model was the one we saw go live within the one year of the project.

In a second step, we developed a strategic assessment of the overall banking economy and the data-driven business models we had created. We also ran an operational and technical gap evaluation: This enabled us to find out where the bank stood in terms of operational excellence and tech maturity.

Building on this, we provided a roadmap to reach the target state, namely to be able to use the new services integrated in the data-driven business model by the time of the go live.

The new AI-driven business model brought the following key outcomes:

  • a platform strategy overview,
  • detailed data monetization options prioritized according to the net business value,
  • implemented open banking system requirements,
  • and an established operating model.

This use case went live within only one year and provided a valuation uplift of 9 million euros within the second year.

Key Takeaways

In today’s constantly evolving environment, banks can leverage data analytics, automation and AI in order to keep abreast of industry trends. For example, they can

  • give personalized financial advice such as product recommendations for a mortgage in the metaverse and thus increase their customer centricity
  • monitor their compliance with ESG requirements which lowers the costs for compliance obligations,
  • automate back-office work to keep down operating expenses

However, to fully realize the potential of data and AI technologies, banks need to do more than just implement individual applications in different parts of their business: they need to make data and AI an integral part of the most important pillars of their business – from strategy to operating model and data foundation.

Your contact


Data analytics Machine learning

7 habits to shorten the time-to-value in process mining
AI ethics AI in business

TechTalk Audio: Responsible AI & ChatGPT
AI ethics AI in business Machine learning

TechTalk Audio: Responsible AI in automotive