Since no target keyword was specified, I’ll use “AI merchant impersonation fraud” as the primary keyword, given the article’s central focus on merchant-side verification gaps in agentic commerce. Category: Fintech. Let me know if you’d prefer a different focus (e.g., “agentic commerce security” or “AI shopping scam risk”).
AI Merchant Impersonation Fraud Exposes Gap in Agentic Commerce Security
The rapid growth of agentic commerce, in which artificial intelligence systems shop and transact on behalf of consumers, has prompted extensive industry investment in verifying that AI agents are authorized to act for legitimate buyers. According to Chris Jones, managing director at payments advisory firm PSE Consulting, a separate and less-addressed risk has emerged: whether the merchants recommended by these AI systems can themselves be trusted.
Testing conducted by UK-based scam-detection service Ask Silver found that fraudulent websites impersonating established British retail brands were appearing within ChatGPT shopping recommendations. Consumers who followed AI-generated purchase links in some cases believed they were transacting with a known retailer, when the destination was a scam site collecting payment credentials.
The issue is relevant to payment networks, merchant acquirers, AI platform operators, and consumers increasingly relying on AI-generated shopping recommendations, particularly in high-velocity retail categories such as fashion, footwear, and electronics.
What Is AI Merchant Impersonation and How Does It Occur
AI merchant impersonation refers to fraudulent websites designed to mimic legitimate retail brands and appear as trusted recommendations within AI shopping assistants, rather than through traditional search engine results. According to the Ask Silver investigation, when ChatGPT was prompted for product recommendations, the AI returned pricing and product details alongside transactional links, some of which routed to scam sites impersonating the retailer Russell & Bromley rather than the brand’s official channels.
The vulnerability was reportedly amplified by circumstances specific to the brand: Russell & Bromley entered administration in early 2026 and was subsequently absorbed into another retail group, creating ambiguity in its digital footprint that fraudsters were able to exploit.
How Fraudulent Merchants Are Surfaced by AI Platforms
Security researchers cited in the report indicate that bad actors are increasingly engineering fraudulent pages specifically to be indexed and surfaced by generative AI tools, rather than optimizing for conventional search engine rankings. Separately, Visa has reported an increase in dark-web forum discussion of AI-agent tools alongside a rise in bot-driven malicious transactions.
Current Buyer-Side Verification Infrastructure
Payment networks have developed extensive protocols to verify that AI agents hold legitimate authorization from consumers. These include Visa’s Trusted Agent framework, Mastercard’s Agent Pay initiative, and developer tools from American Express, which use cryptographic signatures, credential binding, and consent policies. OpenAI and Stripe have developed agent-to-merchant protocols, and Google’s commerce initiatives, involving Shopify, Walmart, Adyen, and Mastercard, use signed spending mandates to limit agent liability. PayPal has also expanded automated AI transaction acceptance across its existing merchant base.
Absence of Comparable Seller-Side Verification
According to the report, no equivalent infrastructure widely exists to verify merchant legitimacy before an AI system recommends a seller. Some card network initiatives, such as Visa’s agentic merchant directory, address whether a merchant portal is technically equipped to support AI checkout, which the report distinguishes from verifying that the merchant is a legitimate business. AI platforms have relied on established marketplace ecosystems, such as Shopify or Etsy, as an implicit legitimacy signal, a proxy the standalone fraudulent sites identified by Ask Silver were able to bypass.
Key Factors Enabling This Fraud Category
Financial analysts note that the underlying economics of impersonation fraud have shifted. Rather than building a convincing storefront to deceive human visitors, a fraudster now needs only to engineer the specific signals that cause an AI system to rank and recommend a fake store. Chris Jones stated that without a comparable merchant-verification layer, AI commerce risks replicating an unregulated marketplace model across the web, with an AI agent, unlike a human shopper, lacking instinct to flag a suspicious domain before payment is initiated.
Cost, Impact, and Implications for Acquirers and Payment Networks
For merchant acquirers, the report characterizes this as a distinct AI-driven impersonation fraud category rather than an extension of existing retail fraud patterns. Because a single optimized fraudulent page can be surfaced repeatedly across automated AI recommendations, acquirers with exposure to high-velocity consumer sectors, including fashion, footwear, electronics, fast-moving consumer goods, and travel, face disproportionate balance-sheet risk as agentic purchasing scales.
The report recommends that merchant website checks, historically conducted primarily at onboarding, shift toward continuous, automated monitoring, with payment networks contributing real-time risk signals drawn from transaction histories.
Risks and Limitations
The report is based on findings from a single scam-detection investigation and statements from one payments advisory source; independent verification of the scale of AI merchant impersonation fraud across platforms beyond ChatGPT is not detailed. Visa’s data on dark-web discussion of AI-agent tools is described in general terms, without specific volume figures disclosed in the report.
No universal global compliance standard currently governs merchant verification within AI commerce, and the report notes that the industry cannot rely on such a standard emerging in the near term. The effectiveness of proposed continuous-monitoring approaches has not yet been independently tested at scale.
Outlook for Merchant Verification in Agentic Commerce
The report indicates that closing this gap will require collaborative data sharing between acquirers, card networks, and AI platform providers to establish shared standards for merchant trust. AI platforms would need to evolve beyond verifying product catalogue feeds toward authenticating full merchant identity. No specific timeline for industry-wide adoption of such standards is disclosed.
Conclusion
The emergence of AI merchant impersonation fraud highlights a structural gap in agentic commerce infrastructure: while payment networks have built extensive systems to verify AI buyer agents, comparable verification of AI-recommended merchants remains largely absent. The longer-term resolution of this gap will depend on coordinated action between acquirers, card networks, and AI platform operators.
