AI in Banking: Trends, Challenges, and What Lies Ahead

In just 70 years, we've gone from lofty bank branches to managing our accounts with a tap on our phones. The banking industry has come a long way in the past few decades. 

Today, we're on the brink of another big shift. As an industry driven by numbers and data, the banking sector is rapidly embracing AI to hit business goals, streamline operations, generate new product ideas, and explore new business models.

By the end of 2030, AI is expected to drive over 1 trillion USD dollars in business value in banking and related financial services, according to Autonomous.

This figure underscores the immense impact AI is having on the financial sector. AI’s influence goes beyond mere efficiency gains – it's transforming banking business models, leading to the emergence of new financial products, services, and even entirely new types of banking institutions (e.g. Neobanks, Banking-as-a-Service (BaaS) Providers, Super Apps).

Advances in data availability, computational power, and algorithms have all made AI more accessible and effective for financial institutions. Those that do not integrate AI into their offerings, operations, and strategy may find it challenging to keep pace with the evolving industry and meet the changing expectations of their customers.

Fraud detection is, without a doubt, one of the top applications of AI by banks. It works by running a set of algorithms that monitor incoming data and prevent fraud before it occurs. Unlike traditional fraud software, AI gets better over time as it learns from the data it processes, adjusting its rules to identify new threats it hasn't encountered before.

We’re also seeing widespread and heavy adoption of AI in areas such as customer service and risk management.

In risk management, AI is transforming how banks assess and mitigate potential threats. By analyzing large amounts of data in real time, AI models can identify patterns and anomalies that might signal financial risks, such as credit defaults or market fluctuations. This allows banks to make more informed decisions and respond fast to emerging risks, reducing potential losses.

When it comes to the front office, AI is completely changing customer relationships and interactions by enabling banks to offer highly personalized financial services to their customers. Through advanced analytics, AI can recommend tailored financial products like loans, insurance, or investment opportunities. This helps banks move beyond basic demographic segmentation and start treating each customer as a unique individual, offering more timely and accurate advice and offers. 

Bank of America, for instance, launched its AI-powered virtual assistant Erica back in 2018 and has since helped with over 32 million client interactions. Fueled by AI and NLP,  these chatbots can handle various requests, from providing updates on balances and initiating transfers to more complex ones such as resolving disputes. 

Digital Transformation and Technology Integration

Not too long ago, banking meant long lines, lots of paperwork, and slow processes. Today, it’s a different story.  While the digital shift began years ago, the pandemic and resulting economic challenges have made it a pressing need. Banks are not only upgrading their digital channels but also integrating advanced technologies such as cloud computing, AI, and blockchain to improve efficiency and enhance customer experiences.

Take JPMorgan Chase, for example. Their AI-powered Contract Intelligence (COiN) platform reviews documents in seconds, a task that would take lawyers thousands of hours. This efficiency is becoming critical as banks look to cut costs and improve service.

The Rise of Fintech and Open Banking

The rise of fintech companies has significantly shaken up the banking industry. These agile and innovative firms are providing alternative financial services that challenge the conventional banking models, especially in areas like payments, lending, and personal finance management. Open banking regulations (e.g. PSD2 in the EU), which mandate that banks share customer data with third-party providers (with the customer's consent), have further accelerated this shift. Open banking is driving competition, promoting collaboration between banks and fintechs, and sparking the creation of new financial products and services.

Heightened Focus on Customer Experience and Personalization

Today’s customers, particularly the younger generations, expect more from their banks. They want quick, personalized service that meets their individual needs, and they want it accessible through their smartphones or laptops. To retain and attract this new wave of customers, traditional banks are increasingly investing in data analytics and AI to deliver personalized experiences.

BBVA is a prime example of a bank that has embraced this shift. They've developed a comprehensive digital sales model that allows nearly 50 million of their customers to interact with the bank online, with 70% of sales happening digitally. BBVA’s onboarding process takes just minutes – much faster than it takes at most banks – by using AI for facial recognition and text analysis to verify applicants through a mobile app.

Increased Productivity Through Generative AI

A recent McKinsey study reveals that GenAI could boost productivity in the banking industry by 2.8% to 4.7% of its annual revenues, translating to an additional value of $200 billion to $340 billion. The significant promise of Generative AI for banks lies in its capacity to enhance the human workforce, especially in bank contact centers.

In the past, banks have depended on the personal touch provided by service representatives to deliver customer service — an approach that, while effective, comes with high costs and limited scalability. Generative AI offers banks an opportunity to scale the provision of high-quality, personalized support to a broader customer base. By integrating GenAI into their operations, banks can boost efficiency and redefine job roles, resulting in a more streamlined and flexible workforce.

Centralized Operating Models for Generative AI

Traditionally, banks have operated with decentralized data practices, where data management and decision-making were often siloed across various departments and regions. However, the unique challenges posed by GenAI, such as hallucinations (the generation of inaccurate information), potential biases, and security vulnerabilities, are driving a significant shift towards more centralized data models. As banks adopt GenAI, they are recognizing the necessity of centralized data frameworks that enable more cohesive and efficient AI deployment. This shift is emerging as a future trend in banking, where unified data management will be key for maximizing potential of AI technologies, ensuring compliance, and maintaining a competitive edge.

Large Language Models (LLMs) in Banking

Due to their ability to analyze large amounts of data in real-time and offer informed, human-like responses that cater to customer’s needs, LLMs are being used for elevating customer experience. LLM can engage in meaningful conversations, making them useful in customer support applications such as chatbots or virtual assistants. 

Though this is the most prominent use of LLMs in banking, they're also used for sentiment detection, portfolio management, and internal and external knowledge management.

One thing is for sure – the industry can anticipate much more innovation and change to follow.

Privacy

The implementation of AI in banking heavily relies on the collection and analysis of large amounts of customer data. This raises serious privacy concerns, especially in light of stringent data protection laws like the General Data Protection Regulation (GDPR) in Europe and the California Consumer Privacy Act (CCPA) in the United States. These regulations require banks to ensure compliance and to be transparent with customers about how their data is being used. 

Regulations

Regulators are increasingly concerned about how AI systems make decisions, particularly in areas like credit scoring and anti-money laundering (AML). The EU AI Act, rolled out in July 2024, classifies AI credit scoring systems as high-risk due to the potential for biased outcomes against individuals or groups. These systems are now required to meet stringent standards for robustness and accuracy.

While these regulations are designed to protect consumers and maintain trust in the financial system, they can also slow down innovation, increase compliance costs, and create uncertainty about how AI can be integrated into banking operations. By proactively incorporating responsible AI practices and collaborating with policymakers and experts, banks can keep driving innovation while maintaining a focus on transparency and accountability.

High Costs

While AI offers transformative potential, the costs associated with its implementation can be prohibitive, especially for smaller banks. AI models (GenAI models in particular) require significant computational power, large-scale data storage, and specialized talent to develop, deploy, and maintain. These high costs can be a barrier to adoption, limiting the ability of some banks to fully leverage AI's capabilities, and potentially widening the gap between large and small players in the industry. Banks need to balance the benefits of advanced AI solutions with the financial and operational resources required to implement them effectively.

Ethical Considerations

Ethical concerns are central to the debate on AI in banking. One of the primary ethical issues is the potential for bias in AI algorithms. These algorithms, often used for decisions like loan approvals, credit scoring, and fraud detection, can unintentionally perpetuate or even amplify existing biases in the data they are trained on. This can lead to unfair treatment of certain groups, such as minorities or lower-income individuals, resulting in unequal access to financial services. 

Biased outcomes threaten trust in AI and create ethical challenges for banks. So banks must design and monitor their AI systems to be fair, transparent, and accountable.

Few technologies have advanced as rapidly or impacted so many industries as deeply as AI. The question of if and when AI will transform the banking industry was settled long ago. The next one is – Are banks ready to take on this opportunity and navigate the challenges that come with it?

The future of banking depends on their ability to do so.

If you want to explore how AI can drive efficiency and innovation in your company, reach out to me at piotr.ho@visium.ch for a conversation.