Answers are grounded in this paper and cite their sources.
## Abstract Agentic commerce moves the buyer out of the audience of advertising. Between September 2025 and September 2026, OpenAI, Google, Amazon, and Meta built protocols and agents that let a model compare products, choose among them, and complete the purchase while the person who wanted the product waits to approve the result. Advertising stayed in the transaction. Amazon made existing sponsored campaigns eligible to appear inside its shopping agent, and OpenAI turned the product feeds it had built for organic results into inputs for its advertising business as well. Merchants meanwhile learned to write for a reader they cannot observe. Research on generative engine optimization found that sources gain visibility in model responses when they add the marks of evidence, such as citations, quotations, and statistics, and the trade literature of 2026 instructs merchants to perfect structured product data for agents. Read as ethnographic material, that literature records a profession redirecting persuasion toward the proxy. The person who delegates a purchase loses sight of the persuasion aimed at it. Advertising disclosure rules assume a human perceiver, and they reach the delegated purchase only where the platform's own recommendation can be treated as a representation to the buyer. Influence that merchants direct at the agent without paying the platform falls outside them, and so does the older law of commercial bribery, which assumes a purchasing agent who can be paid. *Keywords:* agentic commerce, AI agents, generative engine optimization, advertising disclosure, attention economy, trade press, industrial reflexivity, consumer protection
No takes yet. Share an insight, caveat, or question.
Jefferson et al. (2026) studied this question.
Synapse has enriched 8 closely related papers on similar clinical questions. Consider them for comparative context: