Payments, historically seen as a back-office support function that just needed to work, are now being recognised as a business enabler. Today, payment teams have an opportunity to move beyond executing business decisions and play a greater role in shaping them. The reason behind this shift is data – payment teams have more knowledge about the business than almost anyone else. 

'When the data comes together, payments can start steering the business, and that puts your expertise at the centre of how the business grows,' said Gabriel Le Roux, CEO and Co-founder of Primer, a unified payments infrastructure provider, speaking to more than 100 payment leaders from 25 countries at the Primer Showcase 2026 in London. 'The companies that get this right will move faster with greater confidence.' 

Yet despite having masses of payments data, in most companies it remains scattered across providers, systems and teams. 'That doesn’t just slow down payments – it slows down the whole business. Intelligence payments can’t run on fragmented data,' continued Le Roux.  

To solve this issue, Primer has built a single, open AI-powered infrastructure layer across the entire payment lifecycle, from checkout to payout, to provide merchants with more visibility and control. As a unified payments infrastructure, Primer also provides payment service provider (PSP) and fraud partner integration options versus connecting directly.  

The Primer Showcase event covered pressing issues, such as optimisation in the modern payments landscape, cross-border complexity, and where AI is making the biggest impact. It also featured case studies from Airalo, a global eSIM provider, and loveholidays, a large European online travel platform.  

Many speakers emphasised the need to have frictionless, optimal, and secure customer experiences, presenting the right payment method at the right time while at the same time responding to a fast-changing environment. 

'Adaptability and flexibility are a key business driver to maintain relevance and support growth,' said Theo Spyrides, VP Product at Primer. 'As you begin to scale your volumes, optimisation isn’t a nice-to-have – it’s a necessity. Every decision that you make about payments impacts the future of your business. It impacts the speed to launch in a new market, your margins, operational efficiency, and much more.'  

Case studies: Airalo and loveholidays 

Payment processing is a substantial expense for merchants that do business at a large scale, so even a small reduction in cost has a major impact on the bottom line. But, as Spyrides indicated, payments can also enable the next stage of growth for a business.  

In one case study, eSIM marketplace Airalo talked about how working with Primer supported its international expansion. In 2019, Airalo launched its business with one PSP, a single Singapore-based merchant ID (MID). As the eSIM company’s footprint expanded to more than 200 jurisdictions, Primer provided the infrastructure assistance to move quickly to a multi-PSP arrangement, connecting Airalo to three different payment processors.  

Additionally, Primer helped Airalo reduce its payment processing costs by 30% within a year by strategically routing card payments across MIDs in Singapore, the UK and the Netherlands. 'This was a big deal for us because when a cost line item moves by 30%, people take notice,' said Siranjani Sundaram, Head of Payments, Airalo. 'We achieved this primarily by addressing our biggest cost bucket, which is interchange fees, and now we are optimising foreign exchange.' 

The growth needed to happen without diverting engineering resources from Airalo’s core business. It was a straightforward decision to adopt Primer as a payments infrastructure because the cost of development was reduced significantly, according to Tudor Munteanu, Senior Product Lead at Airalo. In addition, Airalo was able to integrate easily with a fraud provider because it was using Primer, as well as activate local payment methods.  

'Primer allowed our product and engineering teams to quickly adapt to the strategies put forward by the finance team to reduce payment costs. It was the tool that helped coordinate multiple departments to make our payments strategy a reality,' he added.  

For Munteanu, using the Primer dashboard and tools such as Observability or AI Companion allows him to easily set up A/B tests, such as testing the cost of processing Brazilian payments. In addition to optimising payment performance, he outlined how the team was able to identify and block a fraud attack in under 20 minutes using AI Companion.  

Another case study featured loveholidays, a fast-growing online travel agency. As loveholidays expanded beyond the UK and Ireland, it had to re-evaluate how to support customers and different payment methods in new European markets with minimal tweaks to its core platform.  

'The North Star for the payments team is to accelerate profitable growth and maximise the value of every transaction, while not being a blocker to customers transacting at the last step of the funnel, or launching a new market,' said Andrea Feinberg, Head of Payments, loveholidays. '[In contrast,] the chief financial officer’s top priority is to keep fees under control.' 

The team frequently reviews whether it makes sense to offer a specific payment method to a cohort of customers in a particular country, and has invested in what it calls a final margin database, which collects revenue and cost metrics across a particular transaction.

loveholidays deployed a multi-acquirer strategy as part of its modern payment infrastructure, as a result of lessons learned during the Covid-19 pandemic when several acquirers pulled out due to a change in risk appetite.  

'We chose orchestration to be able to move quickly between acquirers to continue trading, not for optimising success rates,' explained Feinberg. 'Now we use Primer to understand all the data, such as which acquirer is working better in terms of fees, authorisation rates, checkout, etc. We also invest in working with our acquirers to fully understand what sits behind the data.' 

Identifying AI use cases 

While payment teams are already working on complex data, the biggest change recently is the tooling that can be used to leverage that data. AI, which has long been deployed in fraud management, is now embedded in many tools to collate massive amounts of data and derive actionable insights in a fraction of the time it takes a human.  

For example, Primer’s Autopilot product gives merchants the ability to use AI and machine learning to route payments to the best-performing provider (based on parameters they set). It will be extending this to all areas of the product, including Primer Disputes, to help customers fight disputes automatically. 

Unsurprisingly, AI and agentic commerce dominated discussions during the Primer Showcase, including emerging use cases, as well as concerns and issues.  

When thinking about specific use cases, David Cox, Payments Strategy Lead at UK telecommunications company Virgin Media O2, emphasised the need to take customer preference into account, which may be influenced by demographics. 'If we use AI in a customer-facing product, we need to ensure that customers understand and trust it – and it actually works. AI could prove to be a double-edged sword, similar to social media. It definitely needs good governance and controls,' he said.  

The ability to access and present data back internally to stakeholders at speed has allowed loveholidays’ payments team to open many doors, according to Feinberg. The team is using AI to interrogate more data points than ever before and then translate the outcomes in a clear, nuanced way to different stakeholders.  

She has also been thinking about how agentic commerce could impact booking a holiday by delegating authority to complete the purchase. 'We need to connect that final step in a way that allows us, as a merchant, to continue optimising the architecture we spent years developing. I don’t want to ditch everything to connect to a large language model. I want to still present payment methods in a way that makes sense for that cohort of customers, as well as use the right payment processor, not one that a particular pipe is forcing me to use,' she said.  

In a session discussing where AI fits into a business’s payments roadmap, held under Chatham House Rule, the panellists talked about how AI has materially changed the way payment teams work, the importance of having a global AI strategy and areas where AI is being deployed successfully today.  

'You don’t have to chase the technology, but you do need to chase the use case,' said one panellist from a large European retail chain. 'We’re not only looking at our use of AI to improve performance or data analysis, but also our clients’ use of AI. We need to respond to the new agents clients are deploying.' 

Another speaker reported being an early adopter of key frontier models such as ChatGPT and Claude Desktop. 'We invested a lot in training for staff to build AI fluency. But in terms of productivity, product development and engineering, it’s now the base-case expectation for staff to be proactively learning about these tools,' they said.  

One panellist described building an AI-powered payment regulatory scanner with the in-house legal team. Operating in more than 180 markets, the team was spending a lot of time monitoring regulatory changes. 'It’s both time-consuming and difficult to be proactive. We partnered with our legal team a few weeks ago to map the end-to-end process. Over the next few weeks, we will expand the overall geographic coverage, make it more proactive, and also reduce third-party legal costs,' they explained.   

When thinking about potential risks, many highlighted that AI is also a useful tool for fraudsters. One panellist doubted whether their company, which is not digitally native, is prepared to face the potential risks, including fraud, introduced by AI. 

A panellist from a subscription business raised a specific concern around customers using agents to do recurrent subscription audits, which could lead to a cancellation if judged to be underutilised by the AI agent. 'We have strong daily and monthly active usage metrics, so I don’t think our business is at risk. But it is something that subscription-based businesses will need to watch,' they advised.  

With agentic commerce, some voiced their concern around the possible loss of the customer relationship. A strong merchant or luxury brand may still see direct traffic to their website or store, but agentic commerce could potentially change the overall dynamic of who owns the relationship, the overall billing lifecycle, and the commercial and economic interplay for those that are selling a more commoditised product. 

Overall, the panellists agreed that agentic commerce was not a replacement for existing channels. 'We are looking at this as a potential additional channel. But if we did explore different styles of agentic commerce relevant to our industry, it would have to have a business case of generating revenue or reaching new clientele,' said one speaker.  

Another flagged up Amazon’s decision in September to block Meta’s AI shopping agent, Muse. 'On my to-do list is worrying about whether I will be able to tell whether a transaction is generated by an agent. Muse uses Link (by Stripe), and even if you don’t accept Link on your checkout, it can create a virtual card to mirror human behaviour. Those are the types of things we will be monitoring because the refund cost is borne by the merchant,' they said.  

Despite all the hype around AI, the panellists were sceptical about Gartner’s prediction that 20% of total commerce transactions will be agentic by 2030, arguing that agentic adoption is moving much slower than projected.  

About Joy Macknight