Artificial intelligence is increasingly being harnessed by key companies within cross-border payments.
getty
In the last two years, artificial intelligence has become a consistent focus within the cross-border payments industry. However, while the space was initially filled with a plethora of launches and announcements of new AI-powered products and services, much of the hype and attention surrounding the technology has now shifted elsewhere.
Other technologies including stablecoins have garnered much of the industry’s attention over the last year, particularly as companies increasingly focus on using AI to improve internal processes, with more consumer-facing agentic offerings yet to become more mainstream.
AI-driven efficiency gains have become the key focus
Most of the biggest cross-border payment companies in the world are already actively leveraging AI to improve efficiencies across their entire operations. Money transfer giant Western Union is using AI across its service operations, tech development and marketing teams to achieve productivity gains. Through this usage, the company hopes to reduce the need to hire more staff as AI tools do more of the work and boost productivity.
Remittance player Remitly is taking a similar approach. It has methodically explored where AI can bolster cost efficiency while maintaining or improving productivity across the business. Through this process, Remitly plans to streamline its business, reducing its headcount and operating costs – producing the same or better output with a leaner organization.
“For a growth company like Remitly, AI enables us to build a lot more products faster, while also improving efficiency elsewhere. Ultimately, I’m bringing in all of the skills I can to combine the silos of the past into one very efficient process,” the company’s new CEO Sebastian Gunningham told me during our latest discussion.
During its latest earnings call, Remitly also outlined its expectation that the technology can help reduce the time it takes to build, test and launch new products from years and months to days.
Following Payoneer’s Q1 earnings call, I caught up with CEO John Caplan, who explained that the company has now entered its “second phase of AI”. Payoneer has implemented agentic AI across its platform team to improve efficiency and quality of its output and is now piloting AI agents to bolster its customer service capabilities. The company is now setting its sights on becoming fully AI-native.
“The second phase, which we’re just beginning to enter, is becoming an AI empowered company,” Caplan explained. “What that means is looking at the workflows of our 2,500 people around the world, from operations, compliance, finance, marketing, sales and operations – the entire company – and supporting them with AI to help them do their jobs faster.”
Agentic AI developments pick up pace
Agentic AI – autonomous agents that figure out the steps required and carry them out to complete tasks without requiring human permissions or authentication – is also increasingly becoming an area of focus for companies in the space. The likes of Visa and Mastercard are both already working on enabling these payments with an aim to drive agentic commerce closer to reality.
Visa believes agentic commerce can significantly expand its addressable market because AI agents will likely drive an increase in the number of transactions it processes. Although it remains early days for this type of commerce, Visa says it has already seen growth in agentic commerce and transactions using its agentic tokens.
In March, Banco Santander carried out Europe’s first end-to-end payment completed by an AI agent using Mastercard Agent Pay – another milestone in the development of agentic commerce. Mastercard is now working with OpenAI to enable agent-to-agent payments and has partnered with Google to launch ‘Verifiable Intent’, a new trust layer for agentic commerce which creates a record of what a user authorized when an AI agent acts on their behalf.
In Asia-Pacific, Alipay, the China-based mobile payment app, has launched its new AI wallet, extending its existing AI payment system by enabling users to manage tasks executed by AI agents before and during payments, and analyze their spending after. The company has also developed an Agentic Commerce Trust Protocol to create a standardized language enabling collaboration between AI and service platforms.
Although the mainstream adoption of AI agents remains some way off, these infrastructure developments signal that the biggest cross-border payment companies in the world view agentic commerce as a major opportunity and one that could become more commonplace sooner rather than later. At the same time, AI is helping companies at all levels of the industry improve efficiencies and reduce product time to market, which should speed up the evolution of the space.