The joint release of a white paper by Sunrate and Mastercard at the 2026 World Artificial Intelligence Conference marks a pivotal moment for the global payments industry. Titled Beyond Automation: Defining Agentic Global Payments, the document moves the conversation from incremental process improvements to a fundamental rethinking of how money moves across borders. By positioning agentic AI as the next evolutionary stage after digitization and basic automation, the authors argue that enterprises are on the cusp of a shift where intelligent software entities can autonomously manage entire payment lifecycles. This perspective is not merely theoretical; it is grounded in the concrete infrastructure and real-world use cases that both companies have been developing over the past several years. For treasury and finance leaders navigating an increasingly complex international landscape, the paper offers a roadmap that balances technological ambition with pragmatic governance. It invites organizations to consider not just what AI can do, but how it can be trusted to act on behalf of the business in a regulated, secure, and auditable manner. The timing of the announcement, coinciding with heightened global trade volatility and rising regulatory scrutiny, underscores the urgency of finding smarter, more resilient payment solutions.
At the heart of the white paper lies a clear definition of agentic AI in the payments context: software agents endowed with reasoning, planning, and execution capabilities that can operate within predefined governance boundaries to orchestrate end‑to‑end workflows. Unlike traditional robotic process automation, which typically handles repetitive, rule‑based tasks in isolation, these agents are designed to understand context, make judgments, and coordinate actions across multiple systems, counterparties, and approval chains. This capability enables them to dynamically adapt to changing conditions such as fluctuating foreign exchange rates, evolving compliance requirements, or unexpected disruptions in the supply chain. The paper emphasizes that agentic behavior does not imply uncontrolled autonomy; rather, it is a form of supervised independence where the agent’s decisions are traceable, explainable, and subject to corporate policies. By framing the technology in this way, the authors aim to alleviate fears of runaway AI while highlighting the tangible efficiency gains that can be achieved when intelligent agents are entrusted with complex, multi‑step payment processes.
The document identifies sixteen persistent pain points that continue to hobble B2B cross‑border payments, ranging from fragmented legacy systems and manual data entry to opaque foreign exchange pricing and protracted reconciliation cycles. These challenges are exacerbated by the growing complexity of global trade, which involves navigating a patchwork of regional regulations, diverse banking standards, and varying expectations from suppliers and customers. Traditional automation tools have made incremental strides by speeding up individual tasks such as invoice data capture or payment initiation, but they often fail to address the end‑to‑end nature of the workflow, leaving gaps where human intervention remains necessary. The white paper contends that agentic AI can bridge these gaps by providing a unified intelligence layer that sees the entire payment journey, anticipates bottlenecks, and initiates corrective actions without constant oversight. This holistic view promises to reduce operational friction, lower transaction costs, and improve cash flow predictability for multinational enterprises.
Sunrate and Mastercard bring complementary strengths to the table that make their vision of agentic global payments both credible and actionable. Sunrate contributes its deep expertise in global payment infrastructure, treasury management, and an AI‑native product suite that has been battle‑tested across nineteen‑plus markets. Its platform already supports multi‑currency wallets, virtual commercial cards, and real‑time tracking capabilities that form a solid foundation for embedding intelligent agents. Mastercard, meanwhile, offers a globally recognized payment network, extensive experience in secure transaction processing, and advanced data‑intelligence capabilities that enhance fraud detection and risk management. By combining Sunrate’s operational reach with Mastercard’s trust and security layers, the partnership aims to create a payment ecosystem where AI agents can operate confidently across borders while adhering to the highest standards of data protection and regulatory compliance. This synergy is presented as a blueprint for how fintech innovators and established payment networks can collaborate to drive the next wave of financial innovation.
The white paper goes beyond theory by enumerating thirteen high‑value use cases where agentic AI can deliver measurable impact across the payment lifecycle. These include intelligent supplier onboarding that validates credentials and risk profiles in real time, dynamic payment routing that selects the optimal corridor based on cost, speed, and compliance considerations, and automated foreign exchange management that hedges exposure at the moment of transaction initiation. Other highlighted areas are accounts payable and receivable automation, virtual commercial card issuance, compliance screening against sanctions lists, fraud detection through behavioral analytics, and conversational interfaces that assist treasury teams with query resolution and exception handling. Each use case is illustrated with a narrative showing how an AI agent would perceive inputs, apply reasoning, execute steps, and log outcomes for audit purposes. The emphasis is on end‑to‑end automation that reduces manual handoffs, minimizes errors, and frees finance professionals to focus on strategic analysis rather than transactional minutiae.
To make the concept of agentic AI tangible, the paper walks through a representative workflow that begins with the ingestion of diverse document formats—PDF invoices, scanned purchase orders, and email threads—and ends with fully reconciled books. An AI agent first extracts pertinent data using natural language processing and optical character recognition, validates the information against master vendor records, and conducts real‑time compliance checks against internal policies and external watchlists. Once cleared, the agent determines the most advantageous payment method and currency route, factoring in current FX rates, transaction fees, and settlement timings. It then initiates the payment through the appropriate rails, monitors the transaction status, and automatically matches the settlement advice with the original obligation. Throughout this sequence, the agent logs every decision point, creating an immutable audit trail that satisfies both internal governance and external regulatory requirements. This level of detail demonstrates that agentic AI is not a black box but a transparent orchestrator that can be trusted to act in the company’s best interest.
Trust and governance emerge as recurring themes, with the white paper asserting that technological sophistication alone is insufficient for enterprise adoption. It introduces the concept of Know Your Agent (KYA), a framework analogous to Know Your Customer (KYC) that requires organizations to verify the identity, purpose, and operational boundaries of each AI agent before granting it access to payment functions. Payment tokenisation is recommended to replace sensitive account details with secure, non‑sensitive equivalents that reduce the risk of data exposure. Auditability is ensured through immutable logs of agent actions, decisions, and the data that informed them, enabling both internal reviewers and external auditors to reconstruct the agent’s behavior. Additionally, the paper stresses the importance of cross‑industry interoperability standards that allow agents from different vendors to interact seamlessly within a shared governance model. By laying out these foundations, the authors provide a pragmatic checklist that chief risk officers, chief information officers, and treasury heads can use to evaluate whether an agentic payment solution meets their organization’s security and control expectations.
Sunrate’s AI‑native portfolio is presented as a concrete embodiment of the agentic vision, comprising five specialized agents that target critical pain points in payment and treasury operations. The Payment Agent orchestrates the initiation, routing, and tracking of cross‑border transfers, continuously optimizing for cost and speed. The FX Agent monitors market movements, executes hedges, and provides real‑time rate recommendations to protect margins. The Compliance Agent automates screening against sanctions, politically exposed persons lists, and internal policy rules, flagging anomalies for human review when necessary. The Onboarding Agent streamlines supplier validation by collecting documentation, performing background checks, and updating master records without manual intervention. Finally, the Chat Agent serves as a conversational interface that answers treasury‑related queries, guides users through exception handling, and surfaces actionable insights from payment data. Together, these agents illustrate how modular AI capabilities can be combined to cover the full spectrum of treasury functions while maintaining strict adherence to governance frameworks.
Mastercard’s contribution to the agentic ecosystem is highlighted through its Agent Pay suite, which extends the concept of verifiable intent to machine‑to‑machine and business‑to‑business interactions. Agent Pay enables AI agents to initiate payments on behalf of authenticated entities, with each action cryptographically tied to a verified identity and a clear statement of purpose. The accompanying Verifiable Intent framework ensures that the agent’s motivations are transparent, traceable, and revocable if needed, thereby addressing concerns about unauthorized or erroneous autonomous actions. Mastercard also emphasizes its ongoing work in payment tokenisation, biometric authentication, and real‑time risk scoring, all of which serve to strengthen the security posture of agentic commerce. By anchoring AI activity in established trust mechanisms, Mastercard aims to provide enterprises with the confidence to delegate payment decisions to intelligent agents without sacrificing control or inviting undue risk.
The executive quotes embedded in the white paper reinforce the strategic rationale behind the collaboration. Paul Meng, Co‑founder and CEO of Sunrate, characterizes the mission as making global payments seamless, compliant, and intelligent, suggesting that the true value of agentic AI lies in its ability to embed real‑time, data‑driven insight into every payment decision. He envisions smoother capital flows, reduced operational friction, and a payment function that actively contributes to corporate strategy rather than merely processing transactions. Anouska Ladds, Executive Vice President at Mastercard, offers a complementary perspective, warning that speed without accountability introduces new categories of risk. She stresses that for agentic commerce to scale beyond pilot projects, there must be a clear, auditable chain linking identity, intent, and action. This dual emphasis on empowerment and responsibility captures the central tension that enterprises must navigate as they experiment with autonomous financial technologies.
From a market perspective, the white paper arrives at a time when businesses are under pressure to optimize working capital, mitigate geopolitical risks, and meet rising expectations for environmental, social, and governance (ESG) transparency. Agentic AI offers a lever to improve payment efficiency, which directly impacts cash conversion cycles and reduces the cost of cross‑border trade. Early adopters stand to gain competitive advantages through lower transaction fees, faster settlement times, and fewer errors that can lead to costly disputes or penalties. However, the paper also cautions that successful deployment requires more than buying a software license; it demands investment in change management, staff upskilling, and the establishment of robust AI governance policies. Organizations that treat agentic AI as a strategic initiative—complete with clear success metrics, executive sponsorship, and iterative rollout plans—are more likely to realize its full potential while avoiding pitfalls related to over‑automation or insufficient oversight.
For enterprises looking to explore agentic AI in their payment operations, the white paper offers several actionable steps. Begin by conducting a comprehensive assessment of current payment pain points, identifying areas where manual handoffs, delayed reconciliations, or FX inefficiencies create the greatest cost or risk. Next, pilot a narrowly scoped use case—such as automated supplier onboarding or invoice‑to‑payment matching—using a sandbox environment that allows you to measure performance, validate governance controls, and gather user feedback without exposing production systems. Ensure that any pilot incorporates the Know Your Agent framework, maintains immutable audit logs, and includes clear escalation paths for exceptions. Simultaneously, invest in training your treasury and finance teams to work alongside AI agents, focusing on interpreting agent outputs, managing exceptions, and refining agent parameters. Finally, establish a cross‑functional governance committee that includes representatives from IT, risk, compliance, and finance to oversee the agent’s ongoing performance, approve updates, and ensure alignment with corporate objectives. By following this disciplined approach, businesses can harness the promise of agentic AI while safeguarding the integrity and security of their global payment processes.