Mastercard and Visa Enlist Banks for Agentic Payment Pilots to Accelerate AI-Driven Transaction Automation

Two of the world’s largest payment networks — Mastercard and Visa — have launched coordinated agentic payment pilot programmes with major global banks to explore and validate use cases for AI-powered payments automation. The initiative introduces so-called agentic payments, where artificial intelligence (AI) agents perform context-aware tasks on behalf of consumers and businesses — such as initiating, routing, authorising and completing transactions — aiming to reduce friction, improve security and unlock new efficiencies in the checkout experience.

These pilots bring banks into early testing environments where AI models integrate with core payment systems, risk engines and digital channels to experiment with autonomous payment decisions, enhanced fraud detection, smart payment routing and seamless user experiences. Mastercard and Visa hope that collaborating with financial institutions will both accelerate innovation and help shape emerging standards, regulatory expectations and interoperability frameworks for agentic commerce.

The projects reflect broader trends in digital payments, where AI is rapidly expanding from analytics and fraud scoring into decisioning and automation, blurring lines between payments, banking, and intelligent financial interactions.

Key Highlights

• Agentic payment pilots announced: Mastercard and Visa are enlisting banks to test AI-driven payment workflows.
• Bank participation: Several global and national banks have signed up to integrate agentic capabilities into selected payment channels.
• Use case focus: Context-aware transaction automation, smart routing, real-time decisioning, dynamic authorisation and cross-rail optimisation.
• Strategic goal: Reduce checkout friction, enhance security and expand the role of AI in consumer and commercial payments.
• Collaboration emphasis: Pilot results will guide standards, interoperability and best practices across the wider payments ecosystem.

What Are “Agentic Payments”?

The term agentic payment generally refers to a payments process where AI agents act autonomously or semi-autonomously to assist or carry out payment tasks on behalf of users, rather than requiring every decision to be made manually by a user. In practice, agentic capabilities could include:

• Smart payment routing: Automatically selecting the ideal card, account or payment method based on cost, rewards, credit availability or regulatory requirements.
• Contextual authorisation: Using real-time contextual data — device, behaviour, location and historical preferences — to authenticate and approve payments with minimal friction.
• Autonomous checkout support: Detecting intent and completing transactions without explicit user action, especially in environments like voice commerce, in-app experiences or IoT devices.
• Adaptive fraud and risk management: AI agents that monitor evolving risk patterns and respond dynamically with real-time countermeasures.

Pilot Goals and Why They Matter

1. Improve Consumer and Merchant Experience

Traditional digital checkout can involve multiple steps and decisions. Agentic payments aim to streamline this by reducing clicks, screens and cognitive load — particularly for recurring transactions, subscriptions and high-frequency commerce.

2. Harness AI for Dynamic Decisioning

While AI is already widely used in fraud detection and analytics, these pilots seek to expand its role into real-time payment decisions, where AI models make intelligent choices about routing, settlement method, timing and authorisation based on a mix of factors.

3. Test Safety and Governance Frameworks

Agentic systems operating in payments must balance convenience with governance, auditability, security and compliance. These pilots will help identify control mechanisms, logging standards and ethical guardrails for autonomous payment decisions — crucial for regulatory acceptance and consumer trust.

How Banks Are Participating

Participating banks are integrating Mastercard and Visa’s agentic frameworks into their digital banking platforms, digital wallets and online checkout processes to evaluate aspects such as:

• Risk-aware authorisation: Tightening or relaxing authentication steps based on real-time risk profiles.
• Merchant-agnostic routing: Auto-selecting payment instruments across cards, accounts and tokens to optimise approval rates.
• Context-driven automation: Using user history and preferences to drive more intuitive payment suggestions or completions.
• Fraud mitigation enhancements: Incorporating continuous behavioural analysis and signal aggregation into transaction decisioning.

The pilots are typically scoped within controlled environments, such as selected customer segments, developer sandboxes or partner networks, allowing banks to assess viability before wider rollout.

Why This Could Change Payments

Enhanced User Experience

Agentic capabilities, if proven at scale, could reduce checkout friction to levels comparable with one-tap experiences while adding predictive layers that anticipate user intent — a meaningful upgrade in digital commerce fluidity.

Reduced Declines and Better Routing

Intelligent routing could significantly reduce payment declines, improve authorisation rates and optimise costs for merchants by choosing the best available path across multiple rails in real time.

Better Security With Less Friction

Adaptive authorisation that balances risk and convenience could lead to better security outcomes without forcing additional steps on users, especially in low-risk scenarios.

Blueprint for Future Payments

By developing standards and best practices now, banks and networks are positioning themselves for broader adoption of AI-native payments — potentially influencing regulatory frameworks, interoperability guidelines and cross-border settlement innovations.

Risks and Regulation Considerations

While agentic systems offer promise, they also raise questions:

• Transparency & Consent: Ensuring users understand and opt in to autonomous actions taken on their behalf.
• Auditability: Recording decisions so they can be explained and reviewed.
• Security: Guarding against adversarial attacks that exploit AI logic.
• Regulatory Compliance: Aligning autonomous actions with AML/KYC, PSD2/SCA and local payment regulations.

Mastercard and Visa — along with their bank partners — are working with regulators to address these white-space issues early in the piloting stages.

#AgenticPayments#AIinPayments#EmbeddedAIDigitalPaymentsfintechMastercardTheFinRateVisa