top of page
  • Chesapeake Group

Generative AI in the Banking Sector


Overview:


Generative AI, autonomic systems and privacy-enhancing computation are three technology trends gaining traction in banking and financial services industry vertical.

The increasing adoption of advanced technologies including big data, blockchain, cloud computing, and biometrics generates extensive data. The growing demand to analyze, report, and collect a large volume of data and gain meaningful insights to support banking processes will support the AI market growth in this sector.


Why banks are investing in Generative AI …


Worldwide IT spending by banking and investment services firms was estimated at $623bn in 2022. The largest category of spending was IT services, which includes consulting and managed services and accounts for 42% of total IT spending in the sector at $264bn. The fastest growing category is software, with spending forecast to increase by 11.5% to $149bn.


Gartner predicts that 20% of all test data for consumer-facing use cases will be synthetically generated by 2025. Generative AI learns a digital representation of artifacts from data and generates innovative new creations that are similar to the original but does not repeat it.


Application of generative adversarial networks (GANs) and natural language generation (NLG) can be found in most scenarios for fraud detection, trading prediction, synthetic data generation and risk factor modeling. It has potential because of the ability to take personalization to new heights.


Nvidia is working with Deutsche Bank to implement a wide range of AI and ML applications in the financial sector. The pair will look to use accelerated computing to speed up risk valuations, price discovery, and model back testing. The multi-year partnership will also explore a wide range of applications tailored towards personalized banking, including intelligent avatars.


Potential near-term use cases:


Fraud detection: Training GANs for the purpose of fraud detection produces successful outcomes due to sensitivity being developed after being trained to identify underrepresented transactions. However, employing GANs for fraud detection has the potential to generate inaccurate results, necessitating additional improvement. In 2021, Swedbank, one of Sweden’s largest banks trained GANs using Nvidia GPUs as part of its fraud and money-laundering prevention strategy.


Data privacy: Synthetic data can be used to create shareable data in place of customer data that cannot be shared due to privacy concerns. Further, synthetic customer data are ideal for training ML models to assist banks to determine whether a customer is eligible for a credit or mortgage loan, and how much can be offered.


Risk management: Generative AI is a viable solution for minimizing losses resulting from lack of adequate risk management. It is possible to calculate value-at-risk estimations that show the potential amount of loss in a particular period or to create economic scenarios that are useful for predicting the future of financial markets.


Friendly interactions: Loan decisions are one of the uses of AI. Both decision makers and loan applicants require explanations of AI-based decision-making processes. Conditional GANs are useful to generate friendly denial explanations.



bottom of page