Generative AI in Payments

For many years, the largest financial institutions have applied increasingly sophisticated artificial intelligence and machine learning models to make payment operations, marketing, and risk management more intelligent, efficient, and profitable.  

But, a new type of artificial intelligence (AI), generative AI, burst onto the scene just over a year ago with the unveiling of ChatGPT, followed by other large-language models (LLMs). 

As the name implies, this new AI generates text, images, and other content based on data on which its underlying LLM was trained.

But how can generative AI positively impact payment companies' growth and profit margins? What are payment companies doing with the new technology to solve problems for customers and employees? I seek to deliver answers in this article.

Most banks and credit unions won't have the resources to build large-language models, ceding that ground to the largest banks and fintechs. However, many companies use a much more affordable approach - training existing models, primarily Open AI's ChatGPT, with company data relevant to the use case being implemented or tested. 

With the technology still in its early stages and thus error-prone, banks are justifiably cautious about using generative AI for customer-facing applications. However, payment companies must use Generative AI to make their operations more efficient.

Here are some examples:

1. Personalized Marketing

Card-issuing banks should use data analytics tools to analyze spending patterns and identify attractive customer segments. Next, generative AI would create marketing messages that resonate with individual users in these segments, resulting in incremental purchases. For example, a bank could identify and target frequent travelers to create awareness and usage of the card issuer's travel-related benefits. Additionally, AI can help businesses monitor campaign results in real time, allowing for adjustments and refinements to maximize impact.

Payment plan provider Klarna and several other e-commerce platforms have enabled a personalized online shopping experience by adding a ChatGPT plug-in to their applications. Users receive curated product recommendations, shopping advice, and links to purchase products through Klarna's search and compare tool.

2. Customer Service

Credit and debit cards require more customer care than other consumer and small-business banking products. Customer service is a significant challenge and expense for banks and credit unions that manage their card programs rather than outsource. GenAI can automate time-consuming manual tasks and help customer care agents answer questions.   

Ally Bank partnered with Microsoft to create a GenAI solution that transcribes and summarizes customer service calls, a time-consuming chore previously done manually. SouthState Bank trained an enterprise version of ChatGPT that enables employees to ask about bank policies and procedures, draft emails, and summarize meetings.

3. Risk Management and Compliance

Fraud detection and prevention models have used artificial intelligence concepts like neural networks for many years. Still, the ability of large-language models to ingest and "learn" from massive amounts of data is a game-changer. By detecting elusive patterns and vagaries in vast data sets, generative AI can train itself to identify fraudulent events before they happen. 

A particular type of open-source LLM, a Generative Adversarial Network, or GAN, enables the fraud team to create transaction data that closely mimics actual transaction data. This robust data set helps the system ferret out fraudulent transactions from legitimate ones and block them. 

These examples only scratch the surface of potential use cases for generative AI in payments. Industry participants must start with a comprehensive strategy considering their ability to leverage customer data with AI applications, keep it secure, and comply with data privacy and protection requirements. 

Generative AI will continue to evolve. Your payments business needs to grow with it.  

David Ritter

Financial Services Strategist, CI&T