How Acquirers Can Prepare for Agentic Commerce

What the Report Covers and Why MSPs Should Pay Attention

The 2026 PYMNTS Intelligence report, titled “How Acquirers Prepare for Agentic Commerce: Acquirers in Brazil, the UAE, and U.S.,” draws on survey data collected from 75 acquirers across three countries between January 16 and January 30, 2026. The research was independently designed and produced by PYMNTS Intelligence, with funding from Visa Acceptance Solutions. You can download the report here.

Agentic commerce refers to AI-driven shopping and payment scenarios in which autonomous agents act on behalf of consumers. These agents can browse, compare, build carts, and in some implementations, complete purchases without direct human input at the moment of transaction. This is not a theoretical future state. Acquirers are actively discussing and evaluating these capabilities now, and the ecosystem decisions being made today will shape how dispute liability, fraud risk, and chargeback exposure are distributed across the payment stack for years to come.

For MSPs managing large merchant portfolios, the stakes are particularly high. Anything that reshapes how transactions are initiated, authenticated, and disputed will eventually flow downstream into the chargeback environment you are responsible for managing.

The Infrastructure Gap Is Already Slowing Adoption

One of the report’s most instructive findings concerns the current state of omnichannel commerce, which serves as the baseline from which agentic capabilities will need to scale. Eight in 10 acquirers report being at least somewhat prepared to offer seamless omnichannel shopping experiences. Yet only 32% of the merchants those acquirers work with actually deliver identical payment experiences across digital and physical channels.

The gap between acquirer capability and merchant deployment is driven by cost and infrastructure friction, not by a lack of demand. The report identifies integration expense as the top barrier, cited by 51% of acquirers. Implementation timelines follow closely behind. Fragmented payment infrastructure, legacy system constraints, and limited orchestration capabilities round out the picture.

This matters for agentic commerce because the report is direct about the relationship: agentic commerce assumes a baseline of unified payments capability. Consistent cross-channel experiences, modern fraud tooling, and resilient infrastructure are prerequisites. Where that baseline does not yet exist, the leap to agent-initiated transactions will be difficult to make cleanly.

Most Acquirers Are Waiting, Not Leading

On agentic commerce specifically, the report finds that most acquirers are adopting a conditional posture. Only 23% plan to actively accelerate merchant readiness for agentic commerce in the next three years. The remaining 77% are waiting for signals, whether from consumers, from business clients, or from the lead of large platforms.

Country-level differences are notable. The UAE shows the strongest forward momentum, with 32% of acquirers planning to accelerate merchant readiness. The U.S. sits at 20%, and Brazil at 16%.

The report frames this cautious majority not as passive resistance but as a rational response to unresolved structural questions. Acquirers are correctly identifying that policy alignment and identity redesign must precede technical optimization. It is worth understanding what those structural questions actually are.

The Core Blockers Are Governance, Identity, and Liability

The top internal barrier acquirers cite is fragmented cross-channel payment infrastructure, at 39%. But close behind is a striking figure: 36% of acquirers identify a lack of readiness to support AI agent-initiated commerce as an internal barrier. This is not primarily a technology shortfall. It reflects the absence of clear standards for how agent-initiated transactions should be defined, authenticated, and assigned liability.

The report is consistent on this point across multiple data sets. When asked what changes are necessary to enable agentic commerce, acquirers ranked addressing regulatory uncertainty first, cited by 49%. Establishing identity verification and credential-binding standards for AI agents came in at 43%. Updating authentication rules that currently assume a human user is present followed at 41%. Enhancing fraud-risk scoring to evaluate AI-agent behavior patterns was cited by 40%.

On the liability question specifically, 45% of acquirers say redesigning liability frameworks for agent-initiated transactions is a necessity. That figure rises to 56% among U.S. acquirers. The dispute resolution implication is direct: if liability for agent-initiated transactions is not clearly assigned, every chargeback that arises from one becomes a contested gray area.

Responsibility Is Distributed, But Chargebacks Stay With Merchants

The report maps how acquirers perceive responsibility across the agentic commerce ecosystem. Acquirers see themselves as primarily responsible for fraud prevention, with 40% of surveyed acquirers assigning this function to themselves. Card networks are viewed as the primary custodians of credential security and tokenization. Issuers share responsibility for identity verification and authentication integrity. AI agents are seen as owning the explainability of their own decisions.

Where does responsibility for refunds and chargebacks land? On merchants. The report finds that 60% of acquirers assign refund and chargeback responsibility to merchants or retailers. And that figure holds consistently across all three markets.

For MSPs, this is the critical data point. Whatever governance and identity frameworks eventually emerge to manage agent-initiated transactions, the post-transaction liability exposure will remain concentrated at the merchant level. Your merchants will bear the operational cost of disputes arising from AI-driven purchases, whether or not they had meaningful visibility into the transaction at the time it occurred.

The Four-Phase Roadmap Acquirers Are Following

The report closes with a phased framework for how acquirers can prepare for agentic commerce. Understanding this roadmap helps MSPs anticipate where the market is headed and what will be required at each stage.

Phase one, covering the near term, focuses on policy and identity. This means defining what constitutes an agent-initiated transaction, codifying liability ownership by use case, and standardizing agent identity and credential binding. Nothing downstream can scale until this layer is resolved.

Phase two builds the trust infrastructure. This involves upgrading fraud-scoring models to account for non-human behavior, deploying agent-governance and permission frameworks, and embedding explainability requirements into agent-triggered transactions. The report notes that 96% of acquirers rate agent-governance and permission tooling as important for the next three years.

Phase three addresses interoperable execution. This requires mandating standard orchestration patterns, enabling real-time tokenization across channels, and developing consent and data-governance frameworks. The report notes that 93% of acquirers rate consent and data-governance frameworks as important, even though clear ownership of that function remains unresolved across the ecosystem.

Phase four is where scaled autonomous commerce becomes viable. This involves activating subscription lifecycle management by agents, building reversibility frameworks before expanding agent autonomy into broader use cases, and eventually enabling agents to complete multi-merchant shopping journeys. The report is explicit that this phase only becomes viable once the first three are in place.

The Chargeback Risk Hidden in Plain Sight

The report does not address chargebacks as a primary focus. But the dispute risk embedded in agentic commerce scenarios is not difficult to extrapolate, and it is one of the more underappreciated operational challenges that MSPs will need to prepare for.

Consider a common agentic commerce scenario: a consumer instructs a shopping AI to purchase a specific product the next time it goes on sale. The instruction is given, forgotten, and weeks later the agent executes the order automatically. The product is delivered. The consumer, no longer aware of or interested in the purchase, contacts their bank. The transaction looks unrecognized. A dispute is opened.

This is not a hypothetical edge case. It is a predictable pattern that emerges when the moment of purchase intent is decoupled from the moment of transaction execution. The same dynamic applies to subscription renewals managed by agents, automated cart replenishment, and any other scenario where a consumer delegates purchasing authority to an AI system and then loses situational awareness of what that system has done on their behalf.

The result is a wave of disputes that are not rooted in fraud in the traditional sense, but in confusion, forgotten authorizations, and the absence of clear transaction recognition at the point of delivery or billing. This is structurally similar to the first-party fraud and friendly-fraud patterns that already drive significant dispute volume today, except that the consent layer is even more difficult to document and the transaction chain is more complex to reconstruct.

The liability picture from the PYMNTS Intelligence report reinforces the concern. If merchants carry 60% of chargeback responsibility in the acquirer view, and if agent-initiated purchases generate disputes that merchants have no direct visibility into, the operational burden on merchant portfolios could increase meaningfully as agentic commerce scales. MSPs managing those portfolios at scale will feel that pressure across their entire book.

Preparing Your Merchant Portfolio for What’s Coming

The acquirer roadmap outlined in the report offers a useful parallel framework for MSPs. The near-term priority is understanding where your merchants currently stand on authentication, transaction data capture, and dispute documentation. If your merchants are already operating with fragmented infrastructure and limited fulfillment data, agentic commerce will compound existing vulnerabilities, not introduce new ones in isolation.

The governance investment acquirers are making on identity, explainability, and fraud scoring is worth tracking closely. As these standards develop at the network and acquirer level, MSPs will need to ensure their merchant portfolios are positioned to comply with and benefit from those frameworks. Merchants with integrated transaction and fulfillment data, clear cancellation and authorization records, and automated alert workflows will be better equipped to defend against the dispute patterns agentic commerce is likely to generate.

More immediately, the chargeback management infrastructure your merchants have in place today will matter more as transaction complexity increases. RESOLVE consolidates dispute alerts, including Verifi CDRN, Ethoca Alerts, and Visa RDR, into a single workflow, enabling fast action before disputes escalate into formal chargebacks. DEFLECT addresses transaction confusion at the point of inquiry by sharing enriched order data with issuers and cardholders before disputes are ever filed. And RECOVER provides a structured representment solution for chargebacks that do proceed despite preventive efforts.

As agentic commerce introduces more automated, intent-deferred transactions into your merchant portfolios, each of these solutions addresses a specific point of failure in the dispute lifecycle that is likely to become more frequent, not less.

Ready to Get Your Portfolio Ahead of This Trend?

The PYMNTS Intelligence report makes clear that acquirers are actively working through the governance, identity, and infrastructure questions that will define how agentic commerce scales. MSPs who wait for those frameworks to fully materialize before assessing their own exposure may find themselves playing catch-up on chargeback volume that has already built up across their merchant portfolio. If you would like to discuss how ChargebackHelp’s solutions can help you assess and address chargeback risk across your merchant portfolio as agentic commerce evolves, reach out to our team. We can walk through the current state of your dispute management infrastructure and identify where automation through RESOLVE, DEFLECT, and RECOVER can reduce downstream exposure before it becomes a portfolio-level problem.

Why ChargebackHelp?

ChargebackHelp brings DEFLECT, RESOLVE, and RECOVER together in a single platform designed to automate dispute management across the full transaction lifecycle. For MSPs managing merchant portfolios at scale, this means a coordinated approach to prevention, resolution, and recovery that reduces operational strain, protects merchant account standing, and keeps chargeback ratios within acceptable bounds. As the payment ecosystem moves toward agent-initiated transactions and the dispute patterns that come with them, having an integrated chargeback management partner in place is not a reactive measure. It is a structural advantage.