Axios Live: AI in fintech raises the stakes for consumer trust
Add Axios as your preferred source to
see more of our stories on Google.
MX's Jane Barratt at the Oct. 18 event. Photo: Henri Sagalow for Axios
LAS VEGAS — Financial services companies are racing to adopt AI, but concerns about accuracy, oversight and consumer trust are complicating its rollout, fintech and communications leaders said.
Why it matters: Money and privacy are at stake as the technology is integrated further into transactions and the industry at large.
Axios' Simon Hernandez-Arthur and Ryan Lawler moderated the Oct. 18 roundtable. The event was sponsored by J.P. Morgan.
5 big takeaways from the conversation:
- "It's a very dangerous time," said MX chief advocacy officer Jane Barratt, because people are using anecdotal evidence about what AI can do to jump in like a "drunken sailor … without truly thinking through the guardrails and use cases."
- "You can't be wrong at all" when using AI in financial services, said Shirley Powell, Finastra head of strategic communications. "One of our products does $7 trillion [in] payments a day. What percentage of those are you willing to get wrong?"
- "The proper governance model" can make all the difference, said Amazon global head of institutional payments Nilesh Dusane. "The financial services industry has been using AI for many years — generative AI and agentic AI — and there are certain tasks that can be done" very efficiently, he added.
- Melio chief business officer Prashant Gandhi says the company has three priorities when it comes to AI work: architecture, business and context. The architecture is "the foundation" of the process, the business focus keeps permissions and stakeholders aligned, and context identifies departures from norms that should get human review.
- Consumers may "despise AI," said Fintech Takes founder Alex Johnson of recent polls, "but I have not seen any evidence that that has stopped them from using the product."
Content from the sponsor's remarks:
In opening remarks, Zack Anderson, J.P. Morgan chief data and analytics officer for Payments and Global Banking, said that focusing on what AI gets wrong contributes to people using the technology incorrectly.
- "The real conversation … is where do you delegate? Where do you use [agents]? Where do you not? What kinds of controls do you need?"
