The payments industry is scaling faster than its infrastructure can comfortably handle. Record volumes, proliferating rails, and rapid AI adoption promise greater speed and intelligence—yet beneath the surface, manual processes, integration gaps, and regulatory uncertainty are creating new points of friction.
AutoRek’s recent report, The Future of Payments Operation 2026, frames this moment as a turning point. The question is no longer what payments can do, but how effectively the ecosystem can support its own growth. In a PaymentsJournal Webinar, Nick Botha, Vice President of Payments and Retail Banking at Autorek and James Wester, Co-Head of Payments at Javelin Strategy & Research, discussed the forces reshaping the landscape, the friction building behind the scenes, and what it signals for 2026 and beyond.
The Scalability Squeeze
Scaling payments is a central theme in the report. With global volumes surging across an expanding mix of rails, payments teams are struggling to keep up. Some 69% of respondents cited manual processes and limited automation as their biggest barriers to scalability.
Infrastructure costs and data throughput remain challenges, but people-dependent workflows have emerged as the primary constraint. Decades into digital transformation, a surprisingly high share of processes remain manual.
“I was talking with a relatively large bank about reconciliations,” said Wester. “They were talking about their process problems and why they did things a certain way, which basically ended up being exporting data into a spreadsheet and then printing the spreadsheet. And I said, well, why is that done that way? ‘Diane designed it that way.’ And I was like, well, who is Diane? I’m looking at all the faces on this webinar, and there’s no Diane. ‘Diane actually retired. But we have done it this way because it’s the best way we can figure it out.’
“Nobody had really gone back to look at why the process exists the way it does,” he said. “It was a fine system, but that process has to be automated in some way.”
A Series of Band-Aids
One reason why manual processes remain high is that automation efforts often layer new systems on top of legacy “Band-Aid” fixes—many of which were themselves built to patch earlier workarounds. Rather than simplifying operations, this can entrench complexity.
In other cases, firms attempt to address a single pain point with an external vendor solution. Even when successful, these tools can create additional manual work upstream and downstream. The result is teams that no longer manage the process that was automated, but instead handle surrounding tasks—uploading files, reconciling outputs, or manipulating data to fit disconnected systems.
“The interoperability between your tech stack and the systems and the partners that you work with is so vital,” said Botha. “If something doesn’t fit into your infrastructure, it creates more operational process—the Band-Aid, if you will. It’s just going to create more operational and manual processes.”
What AI Can and Cannot Do
AI adoption, meanwhile, is widespread: 96% of respondents report using AI in some capacity. But there are also some key concerns.Nearly half struggle to integrate AI with legacy infrastructure and are wary of implementation and ongoing maintenance costs. Data security and regulatory risk also rank among the top concerns.
“In the reconciliation world, it’s not just what matches and what doesn’t match, but understanding the behaviors of your clients, settling funds in the right time frames, and helping clients get more value out of certain activities,” said Botha. “But we’ve seen pushback on the macro introduction of AI to take over everything. There’s still a lack of trust in an AI tool to be making decisions on behalf of a firm.”
For now, AI is augmenting rather than replacing human decision-making. It excels at synthesizing quantitative and qualitative data, surfacing insights that enable faster, better-informed decisions. But, it’s not making the final call.
Pattern recognition is where AI delivers the greatest value today. It can process vast datasets and identify signals that would be difficult, if not impossible, for humans to detect.
“Fraud is one of the big ones where if you don’t see that pattern for what fraudsters are doing, they can fly under the radar and exploit it for very large amounts of money across huge amounts of transactions,” Wester said.
AI also drives efficiency in routine tasks. A reconciliation processes that once required four hours of manual transaction matching can now be reduced to minutes.
“You’re talking about not using AI to replace people, but to augment people,” said Wester. “You’re letting that person do things five times quicker than they could, which then gives them the ability to actually contribute to the business. It’s not going to do your job for you. It’s going to allow you to do your job better.”
Looking into the Future
Beyond AI, structural shifts are underway. Central bank digital currencies (CBDCs) and stablecoins have moved from experimentation to strategic consideration. Companies now expect 24% of payment volume to flow through blockchain-based rails by 2030.
The migration to new rails carries regulatory implications as well. Policymakers are still demanding whether certain digital assets should be treated as securities, assets, or payment instruments. One major bank CEO recently warned that if regulators permit yield-bearing stablecoins, it could trigger a migration of as much as $6 trillion in deposits from traditional banking into digital formats.
Finally, market consolidation could accelerate in the next 12 to 18 months. Potential large-scale deals—such as Stripe reportedly exploring an acquisition of PayPal—would reshape competitive dynamics and could amplify the impact of the broader trends already underway.
“Consolidation of the market is something that I’ve been keeping a close eye on,” said Botha. “I’m really interested to see what impact that has on the future of the payments industry.”
