Even though artificial intelligence has come to dominate headlines, product roadmaps and quarterly investor calls, one concept remains stubbornly analog: trust.
“Every major evolution or transformation centers on what matters most: the structure of your business model, what your customers need, expect and trust from you,” Discover® Network Director of Payments Innovation Kate Lybarger told PYMNTS. “And I may argue, trust being the most important of those three.”
That’s particularly true when it comes to AI, she said. While earlier periods of enterprise innovation, such as mobile and digital, were reactions to external constraints, AI introduces internal and existential pressure. Businesses now grapple with not only how to use this new technology, but also how to deploy it in ways that align with evolving customer expectations and regulatory scrutiny.
Lybarger, who has spent her career at the intersection of emerging technologies and enterprise transformation, said AI isn’t a revolution for revolution’s sake. It’s a relatively new tool, albeit a powerful one, that must be wielded with precision, transparency and responsibility.
“If you’re building anything with AI, nothing matters more than trust,” she said. “These solutions require a lot of data, potentially your customers’ information.”
Advertisement: Scroll to Continue
The stakes are at their highest in security-critical industries like financial services, where transactions can be fast, frequent and sensitive.
The New Frontier as Generative AI Is Supplanted by Agentic Commerce
One of the most notable shifts taking place is the transition across payments and commerce from general-purpose generative AI tools to what’s being termed agentic commerce, representing a more targeted, transactional application of advanced AI technologies, Lybarger said.
“We [the industry] are agentic commerce-obsessed right now,” she said, adding that while the technology promises to enhance the customer experience, it’s also important to balance innovation with risk management. Best practices include incorporating extensive testing and human-centered guardrails before introducing AI systems or models into any customer or business setting.
“[There are] all kinds of precedents with technology that has the power to impersonate you,” she said. “The responsible and ethical use of these tools is paramount.”
For Discover Network, this isn’t just a philosophical stance. It’s a strategic imperative. AI is a means, not an end, Lybarger said. It must be applied with caution, clarity and a deep understanding of what problems need to be solved.
“How businesses adopt and deploy AI and how they think about managing situations when things don’t go right, because they inevitably at some point won’t, that matters deeply,” she said.
While one of the more tantalizing promises of AI is its ability to synthesize internal knowledge across departments and eliminate the dreaded corporate silo, AI development has not progressed to that stage yet.
“AI may not be mature enough to be deployed successfully for a given problem at a given time,” Lybarger said. “But when it is, and when it’s applied to the right problem, it can reinforce the value of your business model, perhaps in a way that’s more effective and efficient.”
Building Ethical AI Foundations in Financial Services
Although Lybarger is immersed in AI-related initiatives, she said the technology isn’t the answer for everything.
“Innovation isn’t only about AI,” she said. “Contrary to all the ways in which we can’t stop talking about AI, it is still, and will continue to be, about solving the right problems in the right way, which may or may not require AI.”
Innovation is the “ability to solve a problem in a manner that someone else finds value in,” she said. “In this equation, then I, as a company, deliver value to you, and in exchange, I capture value for me.”
That balanced exchange between delivering and capturing value is where AI finds its rightful place. It’s a supporting actor, not the lead. This perspective can also help temper expectations.
“AI is not good at a lot of stuff yet,” Lybarger said. “Which is hard to believe, given how oversaturated with AI amazingness we are at every turn … [while] AI is solving just about any problem you can think of, the reality is, that shouldn’t be today’s expectation.”
Embracing a level-headed and clear-eyed view of AI’s role in business model evolution can, in turn, raise new organizational questions around deployment and responsibility.
“At the end of the day, people want to make payments, and they want it to work when they need it and in the way they want it to,” she said. “If we can deploy AI solutions to help make that easier, faster, more secure, then we’re going to continue pressure-testing and exploring those opportunities.”
For all PYMNTS AI, B2B and digital transformation coverage, subscribe to the daily AI, B2B and Digital Transformation Newsletters.
