The legal standoff between Amazon and Perplexity AI is shaping up to be the first major test of how artificial intelligence (AI) will reshape the way goods are bought and sold online through agentic commerce.
It began as a dispute over one AI assistant’s behavior. Since then, it has become a case study in who controls the customer relationship when machines — not humans — start doing the shopping. And the implications are profound for B2B ecommerce sellers, from manufacturers and distributors to digital marketplaces. The fight lays bare the tensions between platform control, AI autonomy and data transparency. And it could define the next generation of digital commerce.
Amazon sent a cease-and-desist letter to Perplexity regarding the agentic shopping assistant in Perplexity’s Comet web browser, alleging it violates Amazon’s terms of service by not identifying itself as an agent. The retailer claims that Perplexity’s software “purposely configured its code to not identify the Comet AI agent’s activities,” constituting an “unauthorized intrusion” under the Computer Fraud and Abuse Act.
“AI agents that make purchases on behalf of customers must operate transparently,” Amazon said in the letter, “because transparency protects customers’ data and enables dialogue between service providers and AI developers.”
Perplexity disagrees. The company argues its AI acts on the user’s behalf, inheriting the user’s permission and therefore does not need to label itself as an agent.
In a public response titled “Bullying Is Not Innovation,” Perplexity maintained that Comet functions as a user proxy, not a separate entity requiring disclosure.
“Your AI assistant must be indistinguishable from you,” Perplexity said. “When Comet visits a website, it does so with your credentials, your permissions, and your rights.”
Amazon challenges limits of agentic commerce’s access
At stake is a foundational point of contention for the emerging AI-driven web. Are agentic systems independent actors that must identify themselves? Or are they extensions of the users they represent? The answer is reverberating throughout the ecommerce industry — especially in B2B, where digital transactions are increasingly automated and data-driven.
The dispute may center on a consumer platform, but the precedent it sets reaches deep into B2B ecommerce. As AI systems begin sourcing, comparing and purchasing products automatically, sellers must contend with an environment where the buyer may no longer be human.
For decades, B2B sellers have invested heavily in direct relationships through inside sales, ecommerce portals and personalized digital experiences. Agentic commerce systems threaten to bypass all that. If an AI assistant can evaluate suppliers and place orders autonomously, the seller’s access to the buyer — and the data that sustains it — could vanish behind the machine intermediary.
This is where the Amazon-Perplexity fight hits closest to home for ecommerce teams. Machine-readable product data — not search optimization or email marketing — will increasingly determine visibility. If a company doesn’t structure its catalog, pricing and specifications for AI ingestion, it may never appear in an agent’s procurement workflow.
Benjamin Fabre, CEO and co-founder of cyberfraud management firm DataDome, said the case captures a critical moment for digital sellers.
“The Amazon-Perplexity dispute underscores a fundamental tension in the emerging agentic web: who controls the customer relationship when AI intermediaries act on users’ behalf,” Fabre said. “For brands, this is a wake-up call. They need to think not just about how they appear in search or on social, but how they’re represented — or excluded — by AI agents making decisions for consumers. The brands that win will be those that embrace transparency, structure their data for machine-readable interactions, and build trust with both users and the AI systems guiding their choices.”
How Amazon’s fight could affect B2B ecommerce
For B2B companies, that warning translates directly to product taxonomies, API accessibility and real-time data integration. Those who fail to modernize risk being invisible to the algorithmic layer of commerce.
Amazon’s insistence that AI agents identify themselves signals how large platforms plan to assert control. Similar restrictions could soon appear across B2B procurement hubs like Amazon Business, SAP Ariba and Thomasnet. Sellers dependent on these channels may face new compliance requirements — dictating which AI agents can transact, how they authenticate and what data they can access.
AI commerce happens in milliseconds. Once an agent commits to an order, it expects instant validation, inventory accuracy, and delivery visibility. Sellers whose systems can’t respond at machine speed will lose ground. And if an AI agent misorders or misrepresents data, accountability becomes murky. Amazon’s complaint hints at an emerging gray zone. When an agent acts on behalf of a user, who’s liable when something goes wrong?
The Amazon–Perplexity confrontation is the first visible collision between established ecommerce infrastructure and the coming era of agentic commerce.
Amazon’s position reflects a platform’s instinct to protect its ecosystem and user trust through strict transparency rules. Perplexity’s counterargument reflects a user-centric philosophy: that AI assistants are simply extensions of human intent.
However it ends, the outcome will shape how AI-mediated transactions unfold across industries. For B2B sellers, it’s a signal that commerce itself is entering a new phase. In it, AI agents, not people, decide where to buy, who to trust and how money moves.
Sign up
Sign up for a complimentary subscription to Digital Commerce 360 B2B News. It covers technology and business trends in the growing B2B ecommerce industry. Contact Mark Brohan, senior vice president of B2B and Market Research, at mark@digitalcommerce360.com. Follow him on Twitter @markbrohan. Follow us on LinkedIn, X (formerly Twitter), Facebook and YouTube.
Favorite