GenAI Gives Businesses a Much-Needed Data Assist

Artificial intelligence is hardly a novel concept. According to Mangopay Chief Product Officer Kirk Donohoe manifestations of AI and predictive AI have been part of the technological landscape for quite some time.

However, the evolution of AI, particularly generative AI, has introduced a transformative shift in its approach to data. This evolution is characterized by a focus on unstructured data, marking a departure from the structured data-centric approaches of the past.

“Ninety percent of everything we deal with on a daily basis is unstructured data, and that’s where gen AI has really come to the fore over the last 24 months or so,” Donohoe told PYMNTS in an interview as part of the “What’s Next in Payments: Payments and Gen AI” series.

To elaborate, he explained that instead of providing a data set and asking for a predicted result, as was done historically, generative AI now handles and derives insights from vast unstructured datasets through language queries, yielding results previously unseen in the traditional AI space.

He highlighted strides in fraud prevention as an example of the diverse applications facilitated by harnessing the power of unstructured data with generative AI, particularly the capability to monitor discussions about particular brands across various forums and platforms.

Through a nuanced analysis of communication patterns, generative AI can discern potential attempts to exploit these brands, Donohoe said — a novel approach that goes beyond conventional rule-based workflows and traditional machine learning and enhances fraud prevention strategies.

That impact extends to payment acquiring as well as customer support and success initiatives, where generative AI is streamlining documentation processes and enhancing customer service.

“AI is now scripting the FAQs and helping to put the documentation together for an API,” he noted, locating the relevant answers when someone makes an inquiry.

Embracing Modularity and Flexibility

As the notion of modularity gains ground in the payments sector, with increasingly mature merchants now open to the concept, Donohoe said Mangopay is embracing the trend and adopting a similar modular approach within its operations.

The strategy, which according to Donohoe emphasizes flexibility at both the payments and AI levels, enables merchants to customize their stacks and securely enhance their products.

“Take [for example] a payments product in the Mangopay ecosystem,” he said. “If you want some additional enhancements on the AI side, we will facilitate that in a contained way, protecting your data and helping that product become more intelligent.”

The goal, he added, is to achieve strong interoperability between multiple unstructured data, enabling the AI engine to get “smarter” as the product becomes “more intelligent.”

Regulatory Learning Curve

Asked about the readiness and education of the market regarding the benefits of AI-driven payment solutions, Donohoe acknowledged existing concerns about the unknown aspects of the technology, particularly its potential implications for businesses and data privacy.

Concerns include questions about where data is stored, who has access to it for model training, and the need for clear regulations, an initiative the European Union is pursuing through its landmark AI legislation.

He also acknowledged the legitimate concerns about regulatory frameworks potentially outpacing innovation, while stressing the importance of industry players and governing bodies to foster a regulatory environment conducive to innovation and ethical practices.

Drawing from the experience of Mangopay, a FinTech firm regulated by the Luxembourg Financial Sector Regulator (CSSF) in Europe and the Financial Conduct Authority (FCA) in the United Kingdom, Donohoe highlighted that engaging in these discussions “is the right thing to do” as regulators navigate a learning curve and find their footing in the evolving AI landscape.

For all PYMNTS AI coverage, subscribe to the daily AI Newsletter.