Separating AI Promise from Reality at AFP 2026

As the 2026 AFP conference heads to Las Vegas, treasury and finance leaders are deciding how to manage their time and attention at this year’s event. With thousands of finance professionals expected, the challenge will not be finding technology providers with an AI story. It will be deciding which deserves serious attention, and which don't.

For Gunita Bindra, Head of New Business Development for Paymode at Bottomline, that process uses a simple standard: make providers prove what their technology can do.  

“The first question I would ask [exhibitors] is if their technology is improving,” Bindra recently said. AI is the focal point of AFP this year. Attendees crave information as they place AI bets. Bindra offers up this question for show-goers to use with vendors: “Is this technology learning from the inputs that I’m giving it every day, and getting better?”

“I would ask the payment technology company, whoever I’m speaking with, to prove it to me,” she adds. In other words, go looking for evidence, not terminology. That includes keeping aware of payments systems and networks that may not yet incorporate AI but are well-established and plainly more effective than AI for B2B payments at scale today.

Bindra says attendees will be awash in announcements of AI agents, automation, and increasingly “invisible payments.” So, while AI will unquestionably be the major theme of conversations at AFP 2026, technology is moving so fast that confusion is unavoidable.

The important distinction, Bindra says, is not whether a provider plugs its AI when in pitch mode. It’s whether the capability being discussed is operational and, if so, at what level.

“Almost all companies are saying they're using AI,” she adds. “It’s on us, as users, to figure out if it’s truly AI, if it’s ethical AI, if it is AI that is going to give you return on investment.” If not, you’re likely chasing buzz without making much progress.

For attendees in 2026, that turns a broad AI conversation into a useful one. Is this capability in production, or just on a roadmap? What can the provider demonstrate?

Look Past the Software

AI is making software easier and faster to build. That raises an obvious question for finance leaders: what should a company build itself, and in what cases does a partner create greater value? An expert on network effects in payments, Bindra draws the line there.

“If you think that payment technology is just a bunch of screens, sure, you can build that,” she says. “What you cannot replicate at all is the network, because networks take time.”

She compares it to LinkedIn. “You don't sign up for the LinkedIn software,” Bindra says. “You sign up for LinkedIn because of the people that are already on it.” The same idea applies to payments. Relationships among buyers, suppliers, banks and payment rails take years to build. So do payments expertise and regulatory knowledge.

“You cannot build in payment expertise that a company acquires over decades. You can't quickly switch-on the regulatory compliance knowledge that a network provider gains over years of operation.”

For AFP attendees, that suggests another question: if software becomes easier to reproduce, what does this provider offer that would still be difficult to replicate?

Put 'Time to Revenue' on Your AFP Agenda

Treasury teams often evaluate technology on capabilities, implementation, and cost. Bindra thinks one measure deserves more attention: how quickly the investment produces financial value.

“I would say time to revenue is probably the most important topic, and the most important lever, with which you should be evaluating technology providers,” she says. Her own expectation is aggressive. "As soon as you finish implementation, you should be seeing return on investment the following month," Bindra says.

That forces the conversation to move beyond the demo. What happens after implementation? When does automation begin reducing manual workflows? When does the business begin seeing the return it was promised?

Networks can shorten that path because many connections already exist. Bindra describes core values as automation, security, and monetization, but emphasizes speed.

“The difference maker for [Bottomline's] Paymode is the network effect and the speed at which we achieve it,” she says, illustrating how solution hunters should size up systems.

As AI Agents Move In, Don't Take Humans Out

Bindra also wants finance leaders to think about what AI does with the information available to it, saying: “To me, ethical AI means that I am doing good things for my network on both the buyer and supplier side."

As agentic AI begins making recommendations or taking actions that previously depended on human judgment, the topic is red hot. Efficiency alone is not enough if the result creates a poor customer experience or uses information in ways customers may not like.

“There should still be a human aspect to help tie the bow on AI agents,” Bindra says.

Bring Better Questions to AFP

AFP 2026 takes place November 8–11 at the Mandalay Bay Convention Center in Las Vegas, with AI, payments, fraud, and the future of finance all on the agenda.

For buyers, the goal should be to leave AFP knowing which capabilities are real, differentiated, and can produce measurable value.

Ask if a vendor's AI is operational and get specifics. Ask how it works, and to see its outputs. Ask how quickly the investment will create value. Ask what network, expertise and relationships sit behind any piece of software or point solution. Also, ask what would be difficult to reproduce if software starts getting easier to build. It's a big "if" (but important).

Then return to Bindra’s opening question. What can they prove? That's how to get the most out of meetings and conversations at a pivotal industry event this year, and after.