i2c Wants Payments Executives to Worry Smarter

AI is forcing payments companies to decide how much autonomy they can safely put inside regulated customer interactions.

If companies move too slowly, customer expectations can outrun them, but if they move carelessly, compliance and trust can suffer.

Legacy data silos complicate the shift because AI reaches across fraud, servicing, credit, compliance and customer engagement.

Watch more: What’s Next in Payments With Jason Goldberg of i2c

Payments executives have always had reasons to be paranoid. The trouble now is deciding whether they are worried about the right things.

That distinction sits at the center of i2c’s view of a payments market confronting artificial intelligence, changing customer expectations and technology stacks built before either became a central operating concern. In a What’s Next in Payments interview with PYMNTS, i2c Chief Client Officer Jason Goldberg described two varieties of corporate paranoia, one concerned with surviving what competitors might do and another trained on opportunities that have yet to become obvious.

“‘Survival paranoia’ is defensive,” Goldberg said. “You’re watching for what might kill you. ‘Thriving paranoia’ is offensive. And you’re watching for what you could be the first one to do.”

Three decades ago, Intel Chairman Andy Grove warned of strategic inflection points, those periods when changes in markets or technology alter the assumptions on which companies have been operating. Payments may be living through one now, although some of the industry’s basic rules remain intact, Goldberg said.

Trust, security and reliability still govern payments. What changes around them are products, rails, customer expectations and the technology used to deliver them.

Goldberg said artificial intelligence’s role inside financial institutions is becoming harder to isolate. “Agentic AI has stopped being a nebulous concept,” Goldberg said.

Until recently, much of the conversation concerned efficiency, including automating routine work and lowering the cost of serving customers. Now, he said, it is shifting from “AI assisting people to AI acting on their behalf,” including reasoning and taking action during live, regulated interactions.

That changes what financial institutions must get right. An AI system that reviews a customer service interaction afterward presents one set of risks, while an AI system making or influencing decisions while that customers are still on the line presents another.

Institutions are asking less about how artificial intelligence can take expense out of the business and more about how it can help them compete and grow, Goldberg said. At the same time, AI, fraud, servicing costs and aging infrastructure are becoming difficult to address as separate technology problems because they depend on overlapping systems, data and controls.

The Cost of Mistaking an Inflection Point for an Upgrade

We’d love to be your preferred source for news.

Please add us to your preferred sources list so our news, data and interviews show up in your feed. Thanks!

That is where Goldberg said he sees Grove’s inflection point argument becoming relevant to payments.

“A technology cycle is something you adopt,” Goldberg said. “You buy the module, you check a box, you move on. An inflection point is dramatically different. It changes the architecture of how you compete in the market.”

Goldberg said he places AI in the latter category because it can reach virtually every part of a financial institution, including servicing, fraud prevention, compliance, credit decisions, marketing and cost management. New payment rails or digital assets can be added selectively depending on a company’s business. AI, on the other hand, impacts every function in every business.

That creates a particular problem for institutions operating multiple legacy platforms and systems of record. Fragmentation itself can become the constraint because another layer of technology does not eliminate the underlying divisions among data and workflows, Goldberg said.

For i2c, the response has been to embed artificial intelligence functions and compliance controls within its processing platform. AI-driven fraud detection and servicing, including compliance scoring and real-time fraud detection, are current applications, Goldberg said.

As for timing, Goldberg offered advice for financial services firms.

“If you move into regulated moments without the appropriate guardrails, the compliance controls, the QA, and the auditability, you’ve traded a competitive problem for a regulatory or trust problem,” Goldberg said. “Trust is earned slowly, and it’s lost quickly.”

Waiting carries its own cost. Customer experiences that still appear optional can become expected once enough competitors provide them, Goldberg said. At that point, a company is no longer deciding whether to differentiate. It is explaining why it cannot provide what customers have come to regard as standard.

“The best time to move is while the choice is still yours and before the customer has made it for you,” Goldberg said.

Watch the full conversation with Jason Goldberg for a deeper discussion of:

Jason Goldberg is chief client officer at i2c and has experience spanning payments, lending, and banking across financial services and FinTech.