Trust Receipts: Verifiable Proof for Agentic Commerce
When AI agents buy on behalf of a person or a company, the important question is no longer who clicked. It is who authorized the purchase, what was allowed, and how that proof can be verified later.
Executive summary
A practical explanation of why Trust Receipts matter, how they work, and what merchants gain when identity, authorization, and payment are packaged into a portable proof instead of a platform promise.
Published
2026-06-26
Updated: 2026-06-26
7 min read
Author
Platform Strategy Team
Commerce strategy analysts
The platform strategy team translates AI, commerce, and protocol shifts into actionable guidance for operational teams.
View profileCategory
Agentic Commerce
A decade ago, online buying was easy to describe. A person found a product, added it to a cart, and paid. That flow is changing. More purchases now begin with an AI agent that searches, compares, applies rules, and acts on behalf of a person or a business. That creates a new question for merchants: how do you prove who authorized a purchase when the action came from an agent?

What a Trust Receipt is
A Trust Receipt is a cryptographic receipt for an agent-led transaction. It is not a traditional ticket or a private log entry. It is signed evidence that records who acted, what was authorized, how much was paid, and whether the record can be altered without breaking the signature. The goal is simple: make the proof portable, verifiable, and independent of any single platform.
- 1Who bought? The verified identity of the agent.
- 2What was authorized? The product, rules, and scope of the action.
- 3How much was paid? The amount and the payment rail used.
- 4Can it be changed? Not without invalidating the signature.
The shift is important: trust stops depending on a promise and starts depending on evidence.
Why agent-led purchases are harder to dispute
Human purchases leave behind familiar signals: an account, a card, an email, a session, a history. Agent-led purchases can involve delegated permissions, spending limits, approval rules, multiple platforms, and a payment provider that sits between the buyer and the merchant. If a dispute appears later, those signals are often too fragmented to answer the basic questions with confidence.
- 1Did the agent actually have permission to buy this item?
- 2Was the amount inside the allowed limit?
- 3Did the agent accept these terms on behalf of the buyer?
- 4Which payment rail processed the charge?
- 5Who is responsible if the transaction is questioned later?
A Trust Receipt brings identity, authorization, and payment into one signed record. That means a merchant does not need to reconstruct the story from scattered logs, internal messages, or screenshots. The transaction can be checked against the receipt itself.
How it works
- 1An AI agent starts a purchase in the merchant flow.
- 2The agent identity is verified cryptographically.
- 3The merchant validates what the agent is allowed to buy.
- 4Payment is executed through the selected rail.
- 5A Trust Receipt is emitted and signed.
- 6The receipt is linked into an audit chain.
Each receipt points to the previous one with a cryptographic identifier. That creates a tamper-resistant trail. If a record changes after issuance, the signature no longer validates. The important part is not just that the receipt exists — it is that a third party can verify it without entering the merchant's internal systems.
What merchants gain
- 1Less uncertainty when a purchase is disputed.
- 2Better traceability for automated orders.
- 3Proof of authorization without storing extra sensitive data.
- 4A cleaner path for audit and compliance review.
- 5More confidence from enterprise buyers and agent platforms.
This is not about adding a technical layer for decoration. It is about being able to demonstrate what happened. Once agent volume grows, reputation alone is not enough to support disputes, reviews, or compliance checks.
Why portability matters
A good receipt should not trap the evidence inside one vendor. Trust Receipts are designed to travel with the buyer, the agent, or the merchant. That portability matters because the proof may need to be checked by an auditor, a bank, a buyer, or a different commerce platform later on. When the receipt is portable, verification becomes a property of the evidence instead of a property of the database that issued it.
Trust in agentic commerce will be won by systems that can prove what happened. Trust Receipts turn that proof into something portable, signed, and easy to verify.
Frequently asked questions
Is a Trust Receipt the same as a normal receipt?
No. A normal receipt records a purchase for the buyer. A Trust Receipt is evidence for verification: it records identity, authorization, payment, and integrity in a form that can be checked independently.
Do merchants need to keep more sensitive data to use it?
No. The point is to reduce dependence on internal records and avoid storing unnecessary sensitive information. The receipt should carry the proof needed for verification.
Who can verify a Trust Receipt?
Any authorized party that has the receipt and the verification material. That can include the buyer, the merchant, a bank, an auditor, or another platform.
Does this only matter for B2B commerce?
No. It is especially useful in B2B, but the same problem appears in consumer commerce once agents can act with delegated permissions and spending rules.
Sources and references
Related articles
developer-guide
Building Agentic Commerce #3: Trust Scores — How Agents Decide Who to Buy From
When an AI agent evaluates merchants, it doesn't read reviews or recognize logos. It reads trust scores — 12 machine-verifiable signals that determine search ranking, checkout eligibility, and payment friction. Here's how the system works.
developer-guide
Building Agentic Commerce #4: x402 Stablecoin Payments — When Agents Pay in USDC
How AI agents make on-chain USDC payments using the x402 protocol — multi-chain settlement, EIP-712 signatures, and production-ready stablecoin checkout.
Agentic Commerce
ACP vs AP2 vs x402: Complete Guide to Agentic Payment Protocols
Three protocols are shaping how AI agents handle payments. ACP (Stripe/OpenAI) for fiat, AP2 (Google) for cart mandates, and x402 (Coinbase/Cloudflare) for USDC stablecoins. Here's when to use each.