Mastercard is exploring generative AI uses with a specific focus on operational efficiency and coding assistance.  

Purchase, N.Y.-based Mastercard, which is experienced in leveraging machine learning and AI technology, is exploring generative AI, including coding and co-creation with existing workflows, Jess Turner, executive vice president and head of global open banking and API, said at a recent Mastercard roundtable discussion in New York.

“Generative AI is very different because that’s using the internet, it’s creating new things,” Turner said. “It’s not taking a task and automating it or having deep learning around it.” 

For now, Mastercard would not use generative AI for creation but rather for optimizing existing operations, she said.  

Nearly 92% of banking executives say there is high demand for generative AI and 95% say the tech has the potential to transform the industry, according to Google’s New Banking Survey, released today. The report surveyed 350 IT decision-makers or senior-level executives in the United States between Oct. 2 and 11. 

The survey found that banks are integrating generative AI within financial operations in the following ways: 

• Assisting in code creation for applications and software; 
• Summarizing financial information, including reports or prospectuses; and 
• Enhancing chatbots and virtual assistants. 

AI at Mastercard 

Mastercard continues to use AI to process documents and extract information quickly and seamlessly to reduce manual labor, Turner said.  

For example, Mastercard acquired financial data and insights company Finicity in 2020 for $825 million and has been using the platform to expedite mortgage applications with AI, Turner said. The financial institution aims to reduce the mortgage application timeline to 22 days compared to an industry average of 42, she said. 

Approaching gen AI will require Mastercard to build on its existing AI strategy — which includes keeping principles at the forefront and avoiding inserting bias into the program, Turner said. 

“We run [AI] through data sets on all our models multiple times and nothing goes live” until Mastercard is sure that no bias exists in the AI model, Turner said. Mastercard “will continue to play [with AI] but we’ll make sure we’re doing it responsibly,” she said.