In a payments world buzzing with AI promises, Carlo Bruno (VP of Product, Adyen) and Georgios Kolovos (EMEA Payments & FinTech Leader, NVIDIA) delivered a session that cut through the noise.
At the Money20/20 Europe AI Summit, the two industry leaders offered a grounded look at AI’s role in payments not just as a buzzword, but as a hard-won capability grounded in data, security, experimentation, and long-cycle adoption.
Real Talk: Innovation in Payments Is a Slow Game
Both Bruno and Kolovos acknowledged that AI won’t reinvent checkout overnight and that’s not a failure. It’s a reflection of payments’ complexity. “Processing payments is not optional,” said Bruno. “The first requirement is reliability. Innovation must happen without breaking what already works.”
This tension between reliability and experimentation shaped much of the conversation.
AI Use Cases That Matter Even If They’re Not Sexy
Bruno emphasized that the highest-impact use cases are often the least flashy:
AB testing tools to help merchants experiment safely
Productivity gains from LLMs parsing card scheme manuals
Fraud control through hybrid approaches combining biometrics and behavioral signals
What’s crucial, he said, is building the right foundation robust data layers, proper monitoring systems, and company-wide alignment not just excited engineers.
“It’s boring advice. But that’s how you move faster.”
Agentic Commerce? Maybe. But Not Yet.
Kolovos and Bruno both touched on the rising conversation around AI agents autonomous systems making purchases or navigating commerce.
Is this the future of checkout? Possibly.
But Bruno was clear-eyed:
“We don’t know yet whether agents will replace checkout or become an additional channel like Apple Pay didn’t replace cards, but added a layer.”
That uncertainty is why Adyen is focused on foundational capabilities:
– Tokenization
– Authentication (passkeys, biometrics)
– Merchant acceptance infrastructure
“We’re monitoring everything closely but the stack needs to be ready before the shift happens.”
One of the more candid moments came when Bruno referenced concerns raised by customers during closed-door sessions.
“Authorized fraud voucher abuse, resale, refund fraud is rising. These are technically ‘real’ customers using the system, which makes prevention incredibly hard.”
Add to that the rise of generative AI in identity manipulation, and it’s clear that AI’s role in fraud detection and identity management will be one of the most high-stakes battles in the coming year.
Where to Start? Go Back to the Foundations
When asked how companies should begin their AI journey, Bruno didn’t talk about models or partnerships. He talked about plumbing:
Invest in internal alignment not just innovation teams
Build real-time data flows with strong privacy controls
Monitor results continuously, not just at launch
“It’s not glamorous, but it’s what lets you scale the wins.”
Final Take: AI’s Impact Will Be Slow, Compound, and Deep
The future of payments won’t be decided by flashy demos or moonshot press releases. It’ll be decided by the companies who do the quiet, unsexy work: building infrastructure, securing data, and iterating in real-world conditions.
Bruno and Kolovos made it clear the race to own AI in payments isn’t about who moves first. It’s about who builds sustainably.
🎯 Want to see how embedded finance actually works?
Liberis is on the ground at #Money2020Europe this week. Book a meeting with the team to learn how they’re powering real growth for platforms and the sellers who rely on them.
📲 Like what you’re seeing?
SPONSORED BY
Not affiliated with Money20/20
Bobsguide is a ClickZ Media publication