3 minutes
AI and payment acceptance: more approvals, fewer false declines
Shift from static rules to real-time decisions that drive better outcomes.
James Fry
Executive Lead – Head of Enterprise Product
James Fry is an enterprise payments leader with 15+ years’ experience driving product strategy, global expansion and delivery across cards, APMs and payouts, and advancing industry collaboration.
Key points
- Static routing rules cannot adapt to live network conditions. AI-powered routing uses real-time issuer and network signals to select the highest-performing approval path for each transaction, reducing unnecessary declines.
- Fraud management and payment approval are connected. Visibility into what is being blocked – and why – helps merchants avoid declining genuine customers while maintaining strong fraud controls.
- Worldpay, now Global Payments, helps merchants stay visible as agent-driven shopping grows. Partnerships with agentic commerce platforms, including OpenAI and Google, combined with fraud-portal visibility into agent-initiated transactions, position Worldpay to support both the reach and the risk side of agentic commerce.
This article is part of our Innovation Focus series, covering the latest product updates from Worldpay and Global Payments. Explore more at Innovation Focus.
Every declined transaction is a cost. Not just the revenue lost in the moment, but the friction it creates for customers and the manual effort it generates downstream. Most merchants accept a baseline decline rate as part of doing business. They probably shouldn’t.
AI-powered routing changes the calculus. Instead of directing transactions based on fixed rules or cost alone, it reads live signals – issuer behaviour, network performance, risk indicators – and picks the path with the highest likelihood of success. The difference in approval rates could be material.
Why static routing falls short
Traditional routing logic runs on fixed rules and priority routes, built for predictability rather than performance. The problem is that issuer and network conditions change continually - by time of day, card type, transaction amount and geography. A route that performs well on Monday morning may not be optimal by Friday afternoon.
Without real-time feedback, static rules keep sending traffic down the same paths even if authorisation rates dip, and merchants are left to absorb the consequences in the form of preventable soft declines.
What AI-optimised routing does differently
Worldpay's approval-optimised routing uses real-time decisioning to assess each transaction individually. It considers multiple factors simultaneously – the issuer's recent approval behaviour, the risk profile of the transaction, network performance data – and selects the route most likely to succeed.
'Worldpay's approval-optimised routing uses real-time decisioning to assess each transaction individually.'
Alongside this, cost-optimised routing gives US merchants a second lever: the ability to route dynamically toward the lowest-cost network path without sacrificing reliability. Both options run through the same unified infrastructure, so merchants can set their own priority – approval rate, cost efficiency or a combination of both.
Fraud management that complements, not conflicts
Higher approval rates are only useful if the transactions being approved are legitimate. Fraud management and routing optimisation then become linked – and Worldpay's fraud portal now gives enterprise merchants more granular visibility into transaction activity to aid in understanding consumer behaviour further.
Merchants can see exactly what is being blocked and why, down to the transaction level. Additionally, AI-generated rules can be built by describing what you need in plain language; the portal generates the conditions and shows you the estimated impact before you commit. That means fewer genuine customers caught in overly broad rules, and clearer insight into where fraud risk actually sits.
For merchants dealing with AI-driven commerce – where agent-initiated transactions are beginning to appear alongside standard consumer purchases – that visibility matters even more. The fraud portal now surfaces whether a transaction was agent-initiated, which agent was involved, and what contextual data is available, so merchants can build rules that account for this new category of activity.
How agentic commerce changes the acceptance picture
As more consumers use AI agents to browse and compare products, transaction patterns are changing in ways that affect further fraud management. Agent-led orders can arrive faster, follow different session patterns and carry signals that standard transactions don’t. Worldpay's fraud portal already flags agent-initiated activity, giving merchants a starting point for classifying it before volumes grow.
Visibility extends beyond risk. Through partnerships with agentic commerce platforms, including OpenAI and Google, Worldpay helps merchants stay visible in the channels where AI agents are starting to shop, not just where payments complete.
'As agent-driven transaction volumes grow, and the fraud surface shifts, having dispute management that can keeps pace without manual effort becomes part of the same acceptance strategy matters.'
For US-based software platforms, specifically, AI-powered chargeback protection adds another layer, automatically compiling and submitting the most relevant evidence for each dispute case, with recovery tracking available directly in the iQ portal. As agent-driven transaction volumes grow, and the fraud surface shifts, having dispute management that can keeps pace without manual effort becomes part of the same acceptance strategy matters.
The infrastructure behind the acceptance rate
The payments infrastructure decisions that matter most right now are not about any single feature – they are about whether acceptance, fraud management and dispute handling can work together as the transaction mix changes. AI-driven routing and fraud controls are part of the same answer.
Related insights
