If a shopping AI, not a person, clicks "buy" through an API, who's still legally on the hook when something goes wrong?
How does merchant-of-record status persist through API-based agent integrations?
This explores whether the merchant stays the legally responsible seller (the 'merchant of record', the party on the hook for refunds, chargebacks, taxes and consumer-law duties) when an AI agent buys through APIs instead of a human clicking through the merchant's checkout page.
This explores whether the merchant stays the legally responsible seller, the 'merchant of record', when an AI agent buys through an API instead of a person using the store's checkout page. The collection has no note that discusses merchant-of-record status by name, so this answer is pieced together from neighboring arguments. Together they suggest the status survives, but its supports have to be rebuilt in new places.
The clearest threat comes from consumer law. Many of the merchant's legal duties were written for a human looking at a screen: a clearly labeled 'order with obligation to pay' button, disclosures shown before purchase, a visible way to cancel. Can EU consumer law rules survive agent-to-agent API transactions? argues that once agents transact with each other over APIs, these screen-based protections can't be enforced at all, because there's no screen. His fix is to rebuild the same rights in the protocol itself, for example a standard 'withdrawal API' an agent can call to cancel an order. Read this way, the merchant can stay legally responsible, but only if its obligations travel inside the API rather than sitting on a web page nobody visits.
The second support is knowing who is on the other end of the sale. A seller can't accept liability for a transaction if it can't tell a real customer's agent from a scraper or a fraudster. Can cryptographic signatures help merchants identify trusted AI agents? describes cryptographically signed requests that tell the merchant which agent is calling, what the shopper intends, and how they'll pay. Can credential vaulting alone solve merchant-facing agent identification? makes an easy-to-miss point: keeping the shopper's card details safe is a separate problem from the agent identifying itself to the merchant. Solving the first doesn't solve the second. What identity standards do shopping agents actually need? shows what happens when that identity layer is missing: Amazon simply blocked Meta's agent through its terms of service. That is a merchant using its seller's authority bluntly, because it has no shared standard to negotiate through.
The third support is payments. The design idea that keeps coming up is to separate the agent's decisions from the steps that carry legal weight. Should AI agents make payment decisions independently? puts the agent's uncertain decision-making in its own layer and keeps authorization and settlement predictable and legally enforceable. That separation is what lets responsibility for a sale stay fixed even though an AI made the purchase decision. Does agent-initiated commerce break existing payment infrastructure? proposes signed 'mandates' from the shopper to replace the human cues that fraud systems and dispute processes used to rely on. Will agents adopt cards before stablecoins for payments? explains why cards come first: virtual cards keep existing chargeback and acceptance rules in place, and those rules already assume a defined merchant. Do agentic commerce protocols compete or complement each other? treats the competing protocols as layers a merchant stacks, with each one absorbing a different level of fraud risk. In other words, liability gets split across protocols instead of resting on one checkout page.
One point you might not expect: the harder question may be economic rather than legal. Even if the merchant stays legally responsible, Can an AI agent serve both merchant and user interests fairly? argues that an agent paid through merchant referral fees can't be loyal to both sides. So 'who is the seller' can stay clear on paper while 'who the agent actually works for' gets murky. For contract and tax questions specifically, such as whether an agent platform could itself become the merchant of record the way marketplaces sometimes do, the collection is silent.
Sources 9 notes
Busch argues that agent-to-agent API transactions bypass design-based consumer protections like mandatory buttons and disclosure rules. He proposes parallel protocol-layer obligations, such as withdrawal APIs, to preserve substantive rights in a post-GUI transaction environment.
Visa developed Trusted Agent Protocol with Cloudflare to enable merchants to authenticate AI agents through agent-specific cryptographic signatures carrying intent, consumer recognition, and payment data. The protocol builds on HTTP Message Signatures and aims to bridge the gap between AI-assistant adoption and merchant trust.
Secure credential storage is a unilateral consumer-security control that prevents agent exposure of secrets, but it cannot substitute for a visible, retailer-legible self-identification protocol that both the account holder and merchant must agree to separately.
Amazon's block of Meta's Muse reveals that AI agents need standardized ways to identify themselves and prove authorization to retailers—separate from consumer-side credential security. Without such protocols, retailers enforce access through terms-of-service blocks rather than negotiated standards.
An IMF analysis argues that agent-mediated payments must split probabilistic AI decision-making (Layer 1) from deterministic authorization and settlement (Layers 2–3) because core payment infrastructure requires predictability and legal enforceability that probabilistic reasoning cannot guarantee.
Show all 9 sources
McKinsey argues that payment gateways and fraud engines built around explicit human identity and intent face significant strain when autonomous agents initiate transactions. The firm proposes cryptographically signed mandates as a replacement for human-observable signals fraud systems traditionally relied upon.
Fintech commentator Simon Taylor argues agent payments follow a three-stage sequence driven by increasing complexity: virtual cards leverage mature controls and universal acceptance, stablecoin-settled cards solve cross-border settlement delays, and native wallets enable multi-agent permission-less structures. The progression reflects different tools for different jobs rather than competition.
Boboev argues ACP, AP2, UCP, TAP, and Web Bot Auth each solve different layers—application, governance, discovery, and identity—creating a capability ladder merchants stack rather than choose between, with each layer absorbing different levels of fraud risk.
Azhar argues that agents like Meta's Muse, which earn referral fees from merchants like Expedia, face structural conflicts that prevent unbiased recommendations. The incentive to collect fees, not technology, determines which platforms build or block such agents.
Papers this line draws on 8
The research behind the notes this line reads — ranked by how closely each paper relates.
- The agentic commerce opportunity: How AI agents are ushering in a new era for consumers and merchants
- Payments in the Age of AI Agents
- How Agentic AI Will Reshape Payments
- From Product Search to Preference Articulation: The Economics of Agentic Commerce
- Infrastructure for AI Agents
- Whom Do AI Agents Work For? Role Assignment Induces Sponsorship Bias in LLM Recommenders
- Deep Dive: The hidden liability of agentic commerce
- Magentic Marketplace: An Open-Source Environment for Studying Agentic Markets