ANALYSIS: New Threats, Same Rules for Finance Generative AI

Generative AI is poised to revolutionize the finance and banking sectors by automating tasks, enhancing customer experiences, and providing valuable insights for decision-making. Key use cases such as fraud detection, personalized customer experiences, risk assessment, and more showcase the wide-ranging potential of this cutting-edge technology. Real-world examples from Wells Fargo, RBC Capital Markets, and PKO Bank Polski further demonstrate the impact and potential of generative AI in transforming the financial landscape. Chatbots and virtual assistants have become integral in banking, enhancing customer support and engagement by providing automated, 24/7 assistance.

As shown in Figure 1.4a, VC investments in AI start-ups have seen a steep increase in the United States in recent years, and have resumed growth in China after declining in 2019. Multiple mega investments of more than USD 100 million in the Chinese mobility and autonomous vehicles industry – which is capital-intensive – support this finding. Good person/bad person rules were the second most common approach taken by banks (48%); staff augmentation was second for savings/credit unions; and consultants were second for fintechs. About the same proportion of firms said that these staffing challenges led to longer cycle times for alert investigation, directly impacting compliance operations. 70% of banks and NBFIs face capacity challenges in their compliance operations—meaning that many departments that are “staffed adequately” face at least occasional capacity shortfalls. Especially for level one teams, heavy workloads and repetitive and routine processes can lead to employee burnout.

Customer Insights and Behavior Analysis

Major banks, like Captial One and Citigroup, employ AI to automate back-office operations, thereby reducing processing times and errors. This not only enhances the efficiency of banking operations but also frees up human resources for more complex tasks. Automated portfolio management is an illustration of how Enhanced Investment Decisions are used in a real setting. AI-powered platforms investigate huge amounts of market data, economic factors, and past performance to improve portfolio management.

But given its potential, it’s poised to deliver a significant transformation in bank operations over the next several years. The horizon of embedded finance, pushed further by AI, promises a world where finance isn’t just a sector but an integral part of our experiences, incorporated effortlessly into daily lives, decisions, and aspirations. Inclusivity, personalization, and unparalleled user experiences will characterize the future of embedded finance. Moreover, AI’s scalability, especially with advancements like LLMs and GenAI, means it can adapt in tandem as embedded finance grows and diversifies.

AI security risks: Separating hype from reality

Through its ability to analyze vast datasets rapidly, generative AI contributes to more accurate and secure financial transactions, fostering a dynamic and technologically advanced ecosystem for payment services. Issues such as complex risk assessment, slow customer service, and inefficient data processing are prevalent in the financial and banking sectors. ZBrain adeptly tackles these challenges with its specialized flows, which enable straightforward, no-code development of business logic for apps through an easy-to-use interface. It offers various large language models and templates to choose from, streamlining the creation and customization of intelligent applications. The significance of generative AI in financial services lies in its ability to generate synthetic data, automate processes, and provide valuable insights for decision-making. By embracing generative AI, financial institutions can unlock new opportunities, improve efficiency, mitigate risks, and achieve better outcomes in the dynamic and complex world of finance.

Challenges include addressing these ethical concerns, ensuring model interpretability, and navigating regulatory frameworks in the finance sector. Several generative AI models find application in finance, including Variational Autoencoders (VAEs), Generative Adversarial Networks (GANs), Autoregressive Models, and Transformer Models. One advantage of autoregressive models is their interpretability, as the model coefficients provide insights into the historical relationships between variables. However, autoregressive models assume stationarity, meaning that the statistical properties of the data remain constant over time. Therefore, it is important to assess the stationarity of the data and possibly apply transformations or consider more sophisticated models, such as ARIMA, which incorporates differencing to address non-stationarity.

Not less importantly, GenAI augments endpoint self-healing capabilities through advanced pattern recognition, by analyzing data and adapting responses to attack patterns. This autonomy ensures optimized, secure configurations and enhances overall defense mechanisms. Otherwise, it’s worth highlighting that this transformation also ushers in new challenges, as it can supercharge hackers and malicious actors. Generative AI significantly transforms deposit and withdrawal services in banking by introducing efficiency and personalized experiences. In deposit services, generative AI automates account opening procedures, expediting the Know Your Customer (KYC) process and ensuring compliance. By employing sophisticated fraud detection algorithms that scrutinize transaction patterns, it reinforces security measures, promptly identifying and preventing unauthorized activities to safeguard deposited funds.

GANs have emerged as a powerful tool for credit card fraud detection, particularly in handling imbalanced class problems. Compared to other machine learning approaches, GANs offer better performance and robustness due to their ability to understand hidden data structures. Ngwenduna and Mbuvha conducted an empirical study highlighting the effectiveness of GANs and their superiority over other sampling models. They also compared GANs with resampling methods like SMOTE, showing GANs’ superior performance.

Where Commercial Banking and Payments Is Headed in 2024

For example, JPMorgan Chase’s CoiN technology reviews documents and derives data from them much faster than humans can. Banks usually maintain an internal compliance team to deal with these problems, but these processes take a lot more time and require huge investments when done manually. The compliance regulations are also subject to frequent change, and banks need to update their processes and workflows following these regulations constantly. A report by Business Insider suggests that nearly 80% of banks are aware of the potential benefits of AI in banking. Another report by McKinsey suggests the potential of AI in banking and finance would grow as high as $1 trillion.

Artificial intelligence (AI) can distinguish between valid and suspect activity by examining transaction histories and client patterns. This degree of automation increases market efficiency and liquidity while also lowering trading expenses. AI algorithms can analyze news and social media data, enabling traders to react quickly to events that move the market and profit from inefficiencies in the market.

Less than 70 years from the day when the very term Artificial Intelligence came into existence, it’s become an integral part of the most demanding and fast-paced industries. Forward-thinking executive managers and business owners actively explore new AI use in finance and other areas to get a competitive edge on the market. Due to the lack of specific guidance, counsel in charge of compliance must make sure policy and practice keep up with current with cybersecurity threats. A final security issue is that generative AI increases the complexity of the system, and the more system complexity that exists beyond easy human cataloging, the more likely that there’s an unknown risk. Unknown risks can’t be mitigated, but humility about the level of security and vigilance about the possibility of breach will likely go far to providing best efforts. There’s also the potential for the models themselves to be compromised during creation or during an update by “data poisoning.” A generative AI needs millions upon millions of documents to create its understanding of text or images or audio.

Variational Autoencoders (VAEs) are generative AI models that are widely used in the finance sector. VAEs are designed to learn the underlying structure of the input data and generate new samples that closely resemble the original data distribution. In the context of finance, VAEs work by encoding the input financial data into a lower-dimensional latent space representation. The encoded data is then decoded back into the original data space, reconstructing the input data. Gen AI models ingest as training data the information to which they’re exposed, such as transaction or chat information, and then repurpose it. This is a concern in banking, where data security is paramount and compliance with privacy regulations is required.

One of the best examples of AI chatbots for banking apps is Erica, a virtual assistant from the Bank of America. The AI chatbot handles credit card debt reduction and card security updates efficiently, which led Erica to manage over 50 million client requests in 2019. AI is expected to serve as a vehicle for customer-centric services in the finance industry. While the latest state-of-art neural network architecture may be appealing and provide better accuracy, it’s rarely the best tool for the job due to its complex nature. These AI-enabled toolkits look for outliers that demonstrate data bias and remove them from  the data flow.

Banks are experiencing more destructive cyberattacks — those that result in deleted data, damaged hard drives, disrupted network connections or leave some other trail of digital wreckage in their wake. In fact, 63 percent of financial institutions say they’ve experienced an increase in destructive attacks targeting their organizations. Top-rated AI banking software development-related questions we have been asked countless times. Ethical considerations, data privacy concerns, and regulatory implications are some of the challenges that financial institutions need to address when implementing AI.

The following highlights key use cases of this GenAI platform within the Finance and Banking industry. Look for a comprehensive exploration of generative AI’s role in banking in the next issue of Mastercard Signals. If a data pool reflects that a certain demographic has historically received fewer loans, the AI application could take that fact as prescriptive and discriminate against that group. LSome of the latest AI-powered finance discoveries include the XAI or Explainable AI, DRL or Deep Reinforcement Learning, Quantum Computing, and NLP or Natural Language Processing.

Regulatory Compliance is important for banking entities to maintain legal and ethical operations, safeguard customer interests, and come in accordance to the rules. Automation of compliance processes, improved monitoring capabilities, and increased effectiveness in recognizing and resolving compliance issues are all important contributions made by AI. Leveraging the transformative potential of AI and other cybersecurity solutions, companies are not merely safeguarding their operations but also reshaping the future of finance, one secure digital transaction at a time. AI for fintech equips companies with the capacity to detect and analyze threats in real-time. Machine learning algorithms sift through vast datasets to identify anomalous patterns or suspicious behavior, enabling early threat detection. The same technological advancements propelling this industry forward have also opened doors to new and sophisticated forms of cyber threats.

Still, the use of synthetic data may lessen the compliance risk of training AI technologies. Second, the AI system lifecycle helps assess policy considerations and identify the actors involved in each stage of the lifecycle – from planning and design to operation and monitoring. Feedzai’s platform was deployed at the core of the bank’s existing enterprise systems using the bank’s own data centers. This enabled the Feedzai platform to be the central decision engine for the bank’s online customer onboarding process and verify identity, check eligibility, and assess fraud risk in real time. Our experts can assist you in utilizing AI to generate transformational changes because of their knowledge of artificial intelligence and awareness of the particular problems encountered by the banking industry.

It is to produce forecasts and projections relating to financial performance, earnings, expenses, and other financial metrics. Investment Analysis and Portfolio Management refers to the application of AI algorithms and models to the analysis of historical data, market patterns, and other pertinent elements. It is to assess investment potentialities, reach wise conclusions, and maximize asset allocation within a portfolio.

• Investment management is another area witnessing the profound impact of artificial intelligence.
• For example, Bank of America’s virtual assistant Erica recently reached the milestone of over a billion client interactions since launching in 2018, with nearly 1.5 million per day.
• Generative AI has revolutionised how banks approach testing and reporting, giving them more flexibility, reliability and trustworthiness.
• The fact that these institutions are currently evolving from basic digital systems to processes driven by AI signifies yet another massive shift in the industry.
• First off, the caliber and applicability of the data that AI algorithms are trained on determine how accurate they are going to be.

Read more about Secure AI for Finance Organizations here.