Rokt Newsroom
AI transforming e-commerce
Latest data drop generated at 2026-06-12T10:31:56.995+00:00.
Data Drop
Agentic commerce is moving from browsing to transaction-ready assistants
Signals suggest retail AI is shifting from passive search and keyword ads toward personalized, stateful assistants that can build carts and increasingly execute purchases.
The strongest evidence points to agentic commerce, AI commerce readiness, and AI commerce discovery as the main themes, with retailers and platforms optimizing for agents, chat-based lookup, and AI search.
Limitation: This is directional, not definitive. The evidence shows momentum, but not that agentic shopping has become the dominant retail behavior.
Questions worth asking
Question: What changed in e-commerce?
Answer: The available signals point toward assistants becoming more transaction-ready, not just recommendation tools.
Question: What does this mean for retailers?
Answer: Retailers may need to optimize for agents and machine-readable discovery, not only human browsing.
Question: Is this already settled?
Answer: No. The evidence is still early and more directional than definitive.
Merchant readiness looks like a bottleneck
Early evidence points to AI becoming a measurable acquisition channel, but merchants may be constrained by weak product data readiness and the need to make catalogs machine-readable and trustworthy.
The strongest evidence explicitly says merchants’ ability to capitalize on AI is being constrained by product data readiness and catalog trustworthiness for AI agents.
Limitation: The evidence does not quantify how widespread the readiness gap is, so this should be treated as a developing constraint rather than a proven industry-wide failure.
Questions worth asking
Question: Why now?
Answer: AI commerce appears to be moving fast enough that catalog quality and machine readability are becoming practical requirements.
Question: What are merchants missing?
Answer: The evidence suggests data readiness may matter as much as ad strategy if agents are part of the buying path.
Question: Is this mostly a technical issue?
Answer: Not entirely. The signals also point to trust and authorization becoming part of the problem.
Trust infrastructure is emerging as the real gatekeeper
Discussion increasingly centers around identity, authorization, and AI-native fraud controls as the bottlenecks for agentic commerce.
Visa’s Agent Score, Agentic Directory/Registry, and OpenAI partnership are framed as signs that the battleground is shifting from checkout mechanics to trust infrastructure.
Limitation: This appears more directional than definitive. The evidence indicates a growing focus on trust rails, but not that a single standard has won.
Questions worth asking
Question: What is the main bottleneck now?
Answer: The available signals point toward permissioning and verification, not just better shopping experiences.
Question: Why does trust matter here?
Answer: If agents are acting on behalf of users, identity and authorization become core requirements.
Question: What may people be missing?
Answer: Agentic commerce may be gated by trust rails before it is gated by product discovery.
Shopping assistants are becoming more autonomous
A recurring pattern is emerging: assistants are evolving from simple recommendation tools into persistent systems that remember preferences, build carts, and may complete checkout inside the assistant experience.
The emerging evidence cites Amazon and Microsoft/PayPal as turning assistants into transaction engines that can compare options, track prices, trigger purchases, and complete checkout.
Limitation: The evidence is limited to a small set of signals, so this should be read as an early pattern rather than a broad market conclusion.
Questions worth asking
Question: What changed in the assistant experience?
Answer: The signals suggest assistants are moving from advice toward action.
Question: Why does this matter for e-commerce?
Answer: If assistants can complete transactions, they may sit closer to the purchase decision than traditional search.
Question: Is this already widespread?
Answer: The evidence is still thin, so widespread adoption is not established here.
Discovery is becoming more governed, not just more conversational
The evidence suggests AI-mediated commerce is not only about chat-based discovery; it is also becoming a controlled platform surface with explicit rules, liability, and onboarding requirements.
The strongest signals point to standards-based infrastructure and AI search, while OpenAI’s merchant feed terms, product feed specs, and commerce policies indicate formal governance around onboarding.
Limitation: This is a partial view of the market. The evidence supports a governance shift, but not the full shape of platform competition.
Questions worth asking
Question: What changed for merchants?
Answer: Merchants may need to meet platform rules and feed requirements, not just optimize for conversation.
Question: What does this mean for discovery?
Answer: Discovery appears to be moving into a more structured, policy-driven layer.
Question: Is conversation still the main story?
Answer: Conversation matters, but the evidence suggests governance is becoming just as important.
Retailers may face a new access problem
The available signals point toward major retailers increasingly restricting third-party AI agents from accessing their catalogs even as AI commerce grows.
The strongest evidence says Amazon is increasingly restricting third-party AI agents from accessing its catalogs, alongside broader movement toward agentic commerce.
Limitation: This is based on a specific retailer signal and should not be generalized to all commerce platforms.
Questions worth asking
Question: Why does this matter?
Answer: It suggests access to catalogs may become a strategic control point in AI commerce.
Question: Does this slow agentic commerce?
Answer: It may complicate it, but the evidence does not show the overall trend reversing.
Question: What should reporters watch next?
Answer: Whether access, trust, and onboarding become the main competitive layers around AI shopping.
Contradictions / Tensions
Complicating pair
Dominant narrative: Retail AI is shifting from passive search and keyword ads to personalized, stateful, transaction-ready agentic commerce—while major retailers like Amazon increasingly restrict third-party AI agents from accessing their catalogs.
Tension signal: Visa is building Agent Score, an Agentic Directory/Registry, and OpenAI payment partnerships, with PYMNTS framing the battleground as trust infrastructure, identity, authorization, and AI-native fraud controls.
Why it matters: This suggests the main bottleneck may not be catalog access or assistant capability alone, but formal permissioning and verification. That complicates the idea that agentic commerce is primarily about richer shopping experiences; it may be gated by trust rails first.
Complicating pair
Dominant narrative: AI is rapidly emerging as a measurable retail acquisition channel, but merchants’ ability to capitalize on it is being constrained by weak product data readiness and the need to make catalogs machine-readable and trustworthy for AI agents.
Tension signal: Visa’s Agent Score and Agentic Registry, plus its OpenAI partnership, frame the key issue as whether agents and merchants are legitimate and authorized, not just whether product data is clean enough for AI consumption.
Why it matters: If trust and authorization become the primary gating layer, then merchant readiness is not only a data-quality problem. It becomes a compliance and identity problem, which materially changes how retailers need to prepare for AI commerce.
Complicating pair
Dominant narrative: Across Wayfair, Shopware, Meta, and LinkedIn, the common signal is that commerce and brand discovery are shifting toward AI-mediated, conversational, and standards-based infrastructure, where businesses optimize for agents, chat-based lookup, and AI search rather than traditional browsing alone.
Tension signal: OpenAI is formalizing merchant feed terms, product feed specs, and commerce policies, which makes AI commerce a governed merchant onboarding layer rather than an open conversational discovery layer.
Why it matters: This complicates the discovery-first narrative by showing that AI-mediated commerce is also becoming a controlled platform surface with explicit rules, liability, and onboarding requirements. The shift is not just toward conversation; it is toward platform governance.
