Generative AI Use Case Taxonomy, 2024: The Banking Industry

Arthur Yuen, deputy CEO of Hong Kong Monetary Authority, says the territory’s central bank is preparing to open a regulatory sandbox focused on how financial institutions may use generative artificial intelligence. We’re starting to experiment with it to help customers complete service-related tasks, but it could also help them to manage their money, plan for the future and understand what NatWest can do to help them with those goals. For example, how can GenAI be used to help make the handover from Cora to a colleague as slick as possible?

• For a deeper exploration of these valuable insights, we invite you to join the two public stage sessions hosted by NTT DATA at Sibos.
• The first offered basic help and support which was instructional – guiding customers on how to complete tasks.
• In an era where financial institutions are under increasing scrutiny to comply with Anti-Money Laundering (AML) and Bank Secrecy Act (BSA) regulations, leveraging advanced technologies like generative AI presents a significant opportunity.
• You can then look at how GenAI can help you to not only do this in less time and at lower cost, but also better.

These include tokenization, virtual products and digital wallets, electronic transactions, straight-through transaction processing and product accounting, as well as sophisticated cloud-based risk and financial crime detection models. One European neobank, bunq, is already using generative AI to help improve the training speed of its automated transaction monitoring system that detects fraud and money laundering. A sandbox regime allows banks or others to experiment with new business models or capabilities under the promise of supervisory leniency. It’s a way for central banks to keep a close eye on innovation, while giving banks comfort that they won’t be unduly punished when newfangled tools go awry. Once the central bank is satisfied the banks have a strong culture of risk management around what’s being tested, the products are allowed to be fully deployed. While artificial intelligence was already promising profound changes in the traditional banking business model, the latest innovation in the technology—generative AI—portends a multisensory revolution in banking services.

AI, particularly generative models, offers solutions to these priorities by automating complex tasks, providing personalized customer interactions, and analyzing vast amounts of data to detect fraudulent activities. The versatility of LLMs enables their application in diverse areas such as automated report generation, customer service chatbots, and compliance document analysis. Their ability to process natural language and generate contextually relevant outputs makes them ideal for successfully performing tasks that require subjectivity and producing human-like text.

For instance, in financial services, they can generate detailed reports, summarize regulatory documents, and predict potential compliance issues based on historical data patterns. Sovereign funding enables these banks to focus on long-term investments and growth opportunities and many have invested heavily over the past five to seven years in upgrading their technology infrastructure. As a result, more banks in the region have adopted flexible, scalable cloud-native technologies and modular API-enabled product platforms, as well as platform-centric operating models. They do not have mission-critical systems with a large overhang of technology debt and key man risks from a dwindling pool of resources conversant in legacy programming languages such as Common Business Oriented Language (COBOL). This data-centricity has been a reason why banks have been among the most prolific adopters of AI and other digital technologies.

Closer to customers

While GenAI offers several advantages for the banking and FinTech market, it also introduces risks that need to be effectively mitigated, which may have important implications for financial institutions. In a dynamic banking environment, banks are seeking to differentiate themselves and gain a competitive advantage. Generative Artificial Intelligence (GenAI) is transforming the banking sector, providing innovative solutions that optimise efficiency, enhance security, and increase customer satisfaction. Identifying opportunities to modernize infrastructure, enhance data quality and improve data flows is the critical first step.

Finance in the experience age heralds a new era for customers and banks alike, with embedded finance the key to success. AI contributes to IT development by assisting in software development processes, from coding to quality assurance. It also aids in modernizing legacy systems, ensuring they remain robust and capable of supporting advanced AI applications. Financial institutions must develop strategies to manage input sensitivity, ensuring that LLMs produce reliable and consistent outputs in compliance scenarios. By enhancing the robustness and reliability of LLMs, financial institutions can mitigate risks and ensure the effectiveness of their compliance programs.

By implementing mitigation strategies, financial organisations can balance leveraging the benefits of GenAI and maintaining robust cybersecurity measures. This approach will help safeguard customer data, maintain trust, and drive sustainable innovation in the digital banking landscape. GenAI  offers tremendous potential for enhancing efficiency, personalisation, and customer engagement in the banking sector. However, it also introduces new cybersecurity risks that must be carefully managed.

This involves using interpretable models, documenting decision-making processes, and providing clear explanations to stakeholders. In addition, references should be provided to the material that was used for producing outputs. This ‘human-in-the-loop’ understanding is also critical in recognising and managing the risks opened up by GenAI. If data feeds are incomplete or the training, prompting and monitoring aren’t up to scratch, the technology can slip into bias, hallucinations (false answers) or toxicity (harmful language).

Data privacy considerations across geographies

Much has been written about whether generative AI will conform with the familiar technology hype cycle, and if so, whether the Trough of Disillusionment awaits. Fueling much of this debate is the current high cost of deploying, training and using the technology. This group, drawn from various departments within CaixaBank and its technology subsidiary CaixaBank Tech, will spearhead the bank’s efforts to leverage generative AI. This project aims to scale up gen ai in banking and implement AI use cases across the entire banking group, building upon the success of its predecessor, GenIAl. Join us at the EY GCC GenAI Conclave 2024 to hear from industry experts on flagship event for GCC leaders of leading organizations across India, focussed on trends and topics concerning today’s GCCs. Explore the future of AI content and the critical role of digital watermarking in protecting creators’ rights and ensuring content authenticity.

Concurrently, in Singapore, we worked with the Monetary Authority of Singapore as part of the MindForge consortium to develop a whitepaper that examines the risks and opportunities of GenAI for the financial sector. In our corporate call centre, we are using GenAI for call transcription, summarisation, service request generation and knowledge base lookup, reducing the amount of time needed to handle customer requests while improving our response quality. What’s different with the emergence of GenAI is that we now have the ability to process vast amounts of unstructured data. Coupled with our existing capabilities around structured data, we are well placed to sharpen the outcomes of our current AI use cases while enabling a new class of data-driven use cases.

Banks should look at use cases through the lenses of value creation and risk. In the near term, banks should focus on driving forward the highest value potential opportunities while factoring in the level of risk exposure. The portfolio of AI investments should accelerate broader bank strategic objectives while capitalizing on near-term quick wins that offer clear value with minimal risk.

Across industries, staffing shortages force companies to “do more with less,” leveraging their limited resources for maximum efficiency. Financial institutions are certainly not excluded from this struggle, and resource constraints may be even more pressing as some of the largest banks strive to process millions of transactions each day. GenAI’s power to process ChatGPT information and aid decision-making presents an immediate opportunity to automate many of the manual tasks comprising employee workloads. Whether it’s in building better internal processes or serving clients, banks and lenders must find the right way forward that serves their unique organizational needs in a truly diverse financial services landscape.

As such, leveraging AI to support cybersecurity is an area Red Hat works closely with its customers. “We’re starting to help them work with some of these newer AI-based tools,” notes Sasso. This includes AI-based creditworthiness assessments by banks, as well as pricing and risk assessments, meaning banks must comply with heightened requirements for such AI applications. “When it comes to Gen AI, there’s still constant innovation coming across,” Harmon says. “For banks today, it’s about understanding how and where they can best apply Gen AI while making sure they are collaborating with regulators to evolve the regulations in this space.

After the COVID-19 pandemic sent the adoption of virtual agent technology soaring, companies are now discovering how adding generative AI into the mix can pay dividends. Forward-thinking organizations can remove friction from customer self-service experiences across any device or channel, driving up employee productivity and enabling adoption at scale. The banking industry is currently experiencing a lower adoption of Gen AI (87%) compared to other industries (97%) due to stricter control measures to reduce the risk of data leakage. According to a recent report released by Netskope Threat Labs, phishing is one of the most common cybersecurity threats in the banking industry.

AML policies are designed to prevent criminals from disguising illegally obtained funds as legitimate income. Similarly, GFC encompasses a broad set of regulations aimed at ensuring financial institutions operate within the legal standards set by regulatory bodies. You can foun additiona information about ai customer service and artificial intelligence and NLP. Compliance with these regulations is crucial to avoid hefty fines and maintain the trust of stakeholders.

In the past five years, we have scaled our AI capabilities to make it pervasive across all parts of the bank, delivering tangible outcomes of S$370m for DBS in 2023, more than double that of the previous year. We are confident of growing the economic impact of our AI initiatives in the coming years, affording us greater flexibility to navigate through business and economic cycles. For banks and lenders to overcome the current barriers and fully embrace AI, there needs to be a holistic strategy that can be incorporated on an organization-wide level. And while some banks and lenders have made these integrations to varying degrees of success, others are struggling to fully embrace this next technological chapter. She said she reminds those with whom she works to «lean on concepts and frameworks» that they’ve already built. The banks top a list of the largest banks in terms of AI talent, innovation and leadership.

The business case for such deals should be based on a careful assessment of capabilities and with results from initial use cases. Compared with cross-industry averages, banks use GenAI at a higher rate in marketing (47%), IT (39%), sales (36%), finance (35%) and customer service (24%). Beyond the 17% of banking leaders who reported fully implementing GenAI into their business processes, another 43% indicated they are experimenting with the technology at the enterprise level. Six in 10 said they have deployed at least one GenAI use case to date – the highest of any industry. While the human brain is excellent at reacting to immediate information and making decisions, GenAI can take a bird’s-eye view of an entire information landscape to surface insights hidden to the naked eye.

KPMG professionals have helped banks pilot genAI as information extractors to find anomalies within contracts or flag potentially fraudulent transactions. GenAI has also been used to quickly create bits of code that allow legacy systems to interact with new technologies. Another significant challenge is the integration of AI technologies ChatGPT App within existing banking systems. Many banks operate with legacy systems that might not be compatible with new AI frameworks, which can create costly and time-consuming issues. Ultimately, the goal is to harness the power of GenAI responsibly, ensuring that innovation does not come at the cost of security and customer trust.

Transforming Contract Management In Banking And Enterprises With Generative AI

The material published on this page is for information purposes only and should not be regarded as providing any specific advice, or used by consumers to make financial decision. The third generation of Cora involved reusing those same digital journeys from online and mobile banking in different channels like telephony. This meant customers could contact us via their channel of choice – and instead of queuing to speak to a colleague, they could chat with Cora for help instead. Cora is freeing up time for colleagues to have quality conversations with customers in the moments when they really need that care, empathy and consideration.

Banks are no strangers to technological change and disruption, and they have a long tradition of investing heavily to keep pace with their peers and emergent fin-techs. While this has helped reduce some costs, banks have seen little benefit in their cost-to-income ratios. As certain costs have fallen, regulatory burdens have grown, and it has become more expensive to attract and retain customers. It’s also critical to adhere to a framework that establishes guard rails to govern how GenAI is used.

Enterprising fintech innovators are recognizing the potential for generative AI to create compelling new service offerings for their customers. One such case is Asteria, an IBM Business Partner based in Stockholm, Sweden. They teamed with IBM Client Engineering to build Asteria Smart Finance Advisor, a new virtual assistant based on IBM watsonx Assistant, IBM Watson® Discovery and IBM® watsonx.ai™ AI studio. Insurance can be complicated, and customers naturally want things to be as simple as possible when they interact with providers. Generali Poland, which offers comprehensive insurance services, recognized that its customer consultants were spending most of their time repeatedly fielding basic queries and managing straightforward claims and policy changes.

Another 30% pointed to lack of transparency and accountability, a number that’s slightly higher than other industries. Over half (54%) said that using public and proprietary data sets has been, or likely will be, an obstacle to implementing GenAI. And nearly as many (49%) said they are experiencing challenges moving GenAI from conceptual to practical.

Banks enter the era of GenAI

Model benchmarking provides a standardized approach to evaluating AI performance, ensuring that models meet regulatory and operational standards. Documentation involves maintaining detailed records of model development, training, validation, and deployment processes. The summit promises to bring together banking leaders, fintech pioneers, and AI experts who have successfully implemented AI-driven solutions in areas like fraud detection and data enrichment. In the mid- and back-offices, the benefits include tackling some of the labour-intensive pain points that raise costs and tie-up time that could be more valuably used elsewhere. Properly deployed technology can reduce the overall cost of compliance by 30%-50%, for example, with specific benefits in areas ranging from workflows and reporting to data-driven decision-making. The paper suggests that financial institutions should implement specific controls for AI systems, including monitoring protocols and human oversight.

Bank systems are getting more difficult to manage as banks try new technologies. It means that commercial banks must sharpen their pencils when it comes to liquidity, operational resilience, and understanding how such failures impact their customers – who can now shift their funds with just a few clicks on their mobile phones. Yuen pointed to the March, 2023 collapse of three banks in the US (Silicon Valley Bank, Signature Bank and Silvergate Bank), as well as Switzerland’s shutting down Credit Suisse, as harbingers of new risks to financial stability. Speaking at a conference organized by The Asian Banker, Yuen expressed alarm at the bank failures of March 2023, which demonstrated new risks to financial stability arising from digital innovation.

As generative AI is integrated into our everyday lives and workplaces, understanding its practical implications is crucial for banks, payments companies, and fintechs aiming to stay competitive and relevant. Companies like Hummingbird, Reality Defender, Ntropy, and SQream will showcase their AI solutions with real-world examples and practical applications. Chris’ comments are representative of a growing consensus that banks must navigate AI implementation carefully. The view is that AI must be regulated across the board, but especially in such a significant (and sometimes volatile) sector. If you’d like to know more about how GenAI could benefit your bank and how to realise the potential, please feel free to get in touch. As a result, you not only need to make sure the initial data sets and populations are right first time, but also keep prompting, checking and re-prompting the AI as part of a continuous cycle of input and output.

Banks should act and adopt new forms of AI like Gen AI, but it shouldn’t come at the cost of the livelihoods of millions of people or at the risk of building prejudiced systems. The industry in general is still cautious around scaling up GenAI functions in core products, before conducting rigorous security checks and launch of designated modules, he added. “A lot of the banks we talked to are not ready for scalable adoption of GenAI yet, with a lack of adequate data or infrastructure,” he said.

These AI systems can handle a wide array of queries, from account information to complex financial advice. Benchmarking AI models involves rigorous testing against standard datasets to evaluate their performance. Continuous documentation and updating of AI models ensure they remain compliant with regulatory standards and perform consistently over time. LLMs like Granite from IBM, GPT-4 from OpenAI, are designed to intake and generate human-like text based on large datasets. They are employed in various applications, from generating content to making informed decisions, thanks to their ability to detect context and produce coherent responses. The summit will feature discussions on building and scaling an AI factory, as well as key use cases like fraud prevention and customer service.

Indeed, GenAI, with its ability to collect and interpret financial data on a vast scale, could force some of the Arabian Gulf region’s biggest banks to rethink their already costly digital banking strategies. The call to action emphasizes the need for financial institutions to adopt AI technologies proactively, leveraging their potential to enhance compliance and operational efficiency. By embracing AI, financial institutions can improve their ability to meet regulatory demands, deliver superior customer experiences, and drive innovation in their operations. Advanced AI systems such as large language models (LLMs) and machine learning (ML) algorithms are creating new content, insights and solutions tailored for the financial sector. These AI systems can automatically generate financial reports and analyze vast amounts of data to detect fraud.

Ensuring compliance with diverse regulatory requirements is critical when deploying AI solutions that process sensitive financial data. Regulators require financial institutions to implement robust governance frameworks that ensure the ethical use of AI. This includes documenting decision-making processes, conducting regular audits, and maintaining transparency in AI-driven outcomes. Compliance with these regulations involves providing clear explanations of AI model decisions, ensuring data privacy, and implementing safeguards against biases and discriminatory practices.

The assessment allows the Accelerating Insights initiative to take a more role-based approach, with some roles receiving more technical training than others, according to Bangor’s Director of Strategic Initiatives, Sandra Klausmeyer. With $7 billion in assets, Maine-based Bangor Savings Bank is already readying itself for the AI-fueled future by focusing on its employees. A multinational company adopted our AI contract review platform to streamline contract negotiations, allowing it to compare contract terms against the company’s predefined legal policies. This significantly sped up the review process and reduced the time to finalize agreements by 80%.