Worldpay webinar recap: agentic commerce – building for adoption, not autonomy

Raluca Constantinescu, Lead Editor at The Paypers, shares the key takeaways from the webinar hosted together with Worldpay on ‘Agentic Commerce: Building for Adoption, not Autonomy’.

In this webinar, experts Reshmi Suresh, Head of Agentic Commerce at Worldpay, now Global Payments, Michaela Weber, SVP, Product at Commerce (BigCommerce), Shahar Tal, CTO at Justt.ai, and moderator Dwayne Gefferie discussed where the technology stands and what merchants must do to prepare for an increasingly agent-driven future before consumer adoption accelerates.

The rapid rise of agentic AI is beginning to redefine how consumers discover products, evaluate options, and ultimately complete transactions. What was, until recently, a conceptual discussion around AI-assisted shopping is now materialising into tangible use cases, and merchants are under increasing pressure to understand what agentic commerce means in operational terms and how it will influence aspects such as conversion, loyalty, discoverability, or competitive positioning, to name a few.

The path forward is not yet clear. A fragmented landscape of standards, protocols, and interfaces is making it difficult for merchants to decide where to invest. At the same time, the pace of innovation leaves little room for hesitation.

Early indicators suggest that consumer adoption may accelerate faster than many expect. According to data shared by Worldpay in the Agentic Commerce Report, 60% of UK shoppers would use an AI agent to find the lowest prices, 61% would trust it to balance cost and quality, and, by 2030, Brits expect around 7% of their total online purchases to be made via shopping assistants. This shift is already underway, and the implications are significant for merchants. As agents compress the path from discovery to checkout, traditional differentiating levers such as brand visibility or user experience may be redefined or disintermediated – and the key question is which players will successfully adapt their infrastructure, trust mechanisms, and strategic positioning to remain competitive.

Understanding the protocol landscape and merchant readiness

To frame the landscape, Reshmi Suresh outlined what ‘agentic commerce’ currently entails. She distinguished between closed-loop agents, operating within a single merchant ecosystem, and open-loop agents, which are able to operate across multiple merchants. Within open-loop environments, three levels of maturity are emerging:

• advisory agents – agents helping with product discovery, then handing over to the shoppers;
• agent-driven, human-present – agents helping with discovery and being able to initiate purchases, but with users still involved for confirming purchases;
• autonomous agents – agents transacting directly with merchants using encrypted credentials and protocols, with minimal human involvement.

Today, most activity remains concentrated in the first stage, with very limited adoption of autonomous purchasing.

On the infrastructure side, multiple protocols are being developed to enable agent–merchant interactions. These include frameworks focused on discovery, checkout flows, and completing a purchase, as well as payment-specific initiatives from Visa and Mastercard addressing aspects such as tokenization, identity, and trust. What’s vital right now is making sure that these protocols are interoperable. Reshmi Suresh noted that the industry is not quite there yet, but this is the goal.

From a merchant perspective, agentic readiness is less about payments, which largely build on existing rails, and more about data and infrastructure. Michaela Weber emphasised that product discoverability is the most immediate priority as consumer use of LLMs has skyrocketed, especially for researching items. Therefore, merchants must ensure their product catalogue is structured and optimised for large language models, despite ongoing challenges with taxonomy fragmentation across platforms.

Agentic commerce impact on customer ownership and dispute management

As agentic commerce introduces an intermediary between merchants and consumers, it reshapes ownership of the customer relationship, unit economics of each transaction, and post-purchase accountability. Shahar Tal highlighted that, as of today, default liability for disputes remains firmly with the merchant, even when AI agents execute the transactions, which creates new challenges, as most ecommerce experiences today are optimised for humans and to maximise conversion rates, often at the expense of detailed context. While this works for human shoppers, AI agents require richer, more structured information to perform correctly and make accurate decisions. A lack of context can lead to increased friction and, ultimately, higher agentic dispute rates, which is expected to happen as agent-driven transactions scale. So, merchants must start capturing and storing more granular data across the transaction lifecycle, working closely with payment service providers and dispute management partners to ensure they can investigate and resolve issues effectively.

At a strategic level, Michaela Weber addressed one of the most pressing concerns for merchants: who owns the customer? While AI agents may increasingly mediate discovery and even checkout, she argued that merchant websites will remain critical, as customers are still likely to visit brand sites for browsing and seeing product reviews and videos in one place – something that isn’t yet fully replicated in AI-driven interfaces. However, risks to core ecommerce metrics remain. Merchants have historically optimised for basket size, repeat purchases, and lifetime value, but early agentic transactions tend to be single-item and highly price-driven. This raises concerns about a potential ‘race to the bottom’ on price and delivery speed.

In this evolving landscape, maintaining access to customer data and preserving differentiated brand experiences via their website will be critical for merchants who wish to sustain competitiveness and increase basket size, which is where the primary margin comes from.

Fraud detection and risk management

The introduction of AI agents into the transaction flow is also reshaping fraud detection and risk management, but not by making existing systems redundant. According to Reshmi Suresh, traditional fraud signals, such as device data and behavioural patterns, remain relevant as long as human-led transactions continue to coexist with agent-driven ones.

A key challenge, however, lies in distinguishing between genuine mistakes, friendly fraud, and malicious activity. For example, consumers may claim they did not authorise a purchase made by their agent. While merchants can often prove user involvement, the operational burden of managing disputes still sits with them. So, capturing detailed transaction data – such as knowing where it came from, the protocol used, and recorded intent data – becomes critical as autonomous transactions evolve, and emerging concepts like verifiable intent are expected to also play a role in strengthening audit trails and reducing disputes.

At the same time, new fraud vectors are emerging. These include nefarious agents impersonating legitimate users, as well as malicious users leveraging trusted agents to bypass controls or exploit discounts. In response, merchants will need to make sure they have access to the identity and behaviour of the user, as well as expand their risk models to include agent identity verification, checks regarding the entity that is behind the agent (for example, knowing whether the agent is backed by a trusted provider), and access to agent history.

Here, Shahar Tal emphasised that agent-driven commerce could significantly increase dispute complexity. Current systems rely on clear, traceable user journeys, but agent transactions introduce more uniform and less distinguishable patterns, making traditional detection less effective.

It is worth noting that, currently, friendly fraud accounts for around 70% of the dispute volume, and this percentage is expected to grow as agents make it easier for consumers to challenge transactions. At the same time, merchants face rising operational demands, needing to collect and submit more extensive evidence across fragmented protocols. Despite this, over 50% of merchants still handle disputes with internal teams and in a manual fashion. The long-term direction, according to Shahar Tal, points towards automated, machine-to-machine dispute resolution – where structured data replaces today’s fragmented and manual processes.

What merchants should ask their providers and partners today

As agentic commerce evolves, the panel highlighted the importance of asking the right questions across partners and shared the following advice for merchants:

• Prioritise visibility and ensure that commerce platforms can make products discoverable in agentic commerce environments.
• Assess how prepared payment providers are to support new protocols, tokenization models, and fraud signals, as well as their role in influencing frameworks and advocating for merchants to ensure their voices are heard.
• Look beyond high-level promises when evaluating dispute and fraud partners, focusing on tangible technological capabilities and proven readiness to handle increasing complexity.

Looking ahead

The closing message is clear: while agentic commerce is evolving rapidly, merchants should focus on assessing foundational capabilities and their readiness, and they do need to act now. The merchants positioned at the forefront as agentic commerce evolves are those who prioritise data readiness, understand internal capabilities, and proactively address emerging risks.

This webinar recap only highlights the key points of the discussion. For the complete take on what merchants must do to prepare for an increasingly agent-driven future, including the insightful Q&A session at the end, watch the webinar recording here.

About the author

Raluca Constantinescu is Lead Editor at The Paypers, where she shapes editorial strategy and helps connect the payments and ecommerce ecosystem, from merchants and marketplaces to PSPs and fintech leaders. With over seven years of experience, she leads a wide range of initiatives, including flagship industry reports, strategic partnerships, and editorial projects spanning in-depth market analysis and thought leadership. Passionate about the future of digital payments and commerce, Raluca is dedicated to creating meaningful connections and delivering insights that drive growth. Reach her via LinkedIn.

 

About Worldpay, now Global Payments

Worldpay, now Global Payments, powers global commerce for millions of merchants worldwide. Our payments technology processes over 50 billion transactions annually across 174 countries and 135 currencies. We help businesses take, make, and manage payments through global acquiring, cross-border capabilities, intelligent optimisation and deep industry expertise. With Worldpay, now Global Payments, merchants gain a partner committed to unleashing their growth potential.