Mastercard Signals Report Explores Trust Frameworks for the Future of Agentic Commerce

Mastercard’ s latest Signals report “Encoding Trust: The Race for Intent, Consent and Control in the Agentic World” examines what it will take for agentic commerce to scale as AI agents move beyond searching and recommending to making decisions and transacting on behalf of consumers and businesses.

The opportunity is significant. Agent-assisted consumer spending is forecast to reach $3–5 trillion by 2030, with business-to-business spending potentially three times higher. The report shows that growth will be driven not by autonomy alone, but by useful experiences that make everyday decisions easier while keeping people informed and in control.

From assistance to autonomy

Agentic commerce marks a shift from AI that supports decisions to AI that can act within conditions set in advance. For consumers, this can mean less time spent on routine tasks and more support in finding the right option. The findings show that 85% of consumers are open to collaborating with an AI agent to find the best option, while 74% are open to letting it complete specific commerce tasks at their request but only one in ten1 are willing to let an agent complete a purchase autonomously. So, people are ready for this kind of collaboration in case the purpose is clear, and they remain in charge of the boundaries.

A typical experience could be simple: regular purchases and payments stay within the budget a customer has set, while an unexpected price change or renewal is paused and brought back for approval. The routine is handled quietly, the exception is explained clearly and the final choice stays with the customer’s approval.

Encoding trust into every transaction

Mastercard’s Signals Report outlines an emerging trust layer between an agent’s decision and payment execution. Before money moves, this layer can verify who is acting, what the user authorized and which limits apply, while maintaining an auditable record of the agent’s activity.

Trust frameworks can play a central role in this model. Beyond protecting payment credentials, agentic tokens can carry task-specific authority and be restricted by agent, merchant, category, spending limit, timeframe or usage rules. Combined with verifiable intent, they can connect a user’s permission to the resulting transaction in a form that can be checked and enforced.

Building shared confidence and responsibility

For agentic commerce to move confidently into daily life, trust needs to be designed into the experience from the start. Users need to know what an agent is doing, change or stop a task when needed and have a clear path to resolve exceptions. Nearly one in three consumers are already open to an AI agent choosing what to buy based on their preferences and budget, provided they retain final approval of the payment. These safeguards turn control from an added step into part of the service itself.

Responsibility therefore extends well beyond the point of payment. AI platforms, banks, merchants and payment networks need shared approaches to identity, permissioning, liability and dispute resolution. Merchants also need machine-readable product information and permissioned preference signals so agents can evaluate offers accurately while preserving user choice.

Mastercard’s role in trusted delegation

Mastercard is bringing together its capabilities in identity, cybersecurity, payment security and network intelligence to support trusted delegation in agentic commerce. These capabilities can help authenticate agents, validate permissions and enable secure transactions across platforms.

The report concludes that autonomy alone will not define the next phase of commerce. Its success will depend on clear permission models, secure infrastructure and controls that people and institutions can understand, use and trust.