Digital Commerce 360 is reporting a public dispute between Amazon and AI search company Perplexity, a clash the publication frames as the first visible fault line between marketplace operators and AI shopping agents in B2B ecommerce.

The fight centers on catalog access. Perplexity, an AI answer engine, has built shopping and product-comparison features that surface merchant listings to buyers. Amazon has moved to restrict that access, drawing a hard line between organic marketplace traffic that converts inside Amazon's checkout and off-site AI referrals that divert demand before the transaction closes. The dispute has not yet been litigated on the specific facts of B2B catalog data.

Why this matters to distributors

B2B marketplaces have largely cloned the Amazon consumer playbook: open catalog data, third-party sellers, programmatic placement, and take-rate economics. If Amazon blocks AI agents from indexing its catalog, expect the same gatekeeping posture to migrate into industrial and MRO marketplaces operated by Amazon Business, Grainger, MSC Industrial, McMaster-Carr, and a long tail of vertical platforms. Margin in B2B ecommerce depends on take rate, sponsored placement, and data licensing — all of which turn on who controls the catalog and on what terms.

What is actually being contested

Perplexity's position is that AI agents deserve the same access as a browser or a search crawler: read the page, surface the result, and let the merchant keep the conversion economics. Amazon's position, as reported by Digital Commerce 360, is that the catalog, the listing quality, the reviews, and the checkout function as a single product, and that scraping that product to drive a competing interface crosses marketplace terms of service. The disagreement is not new in principle — comparison shopping engines fought the same fight in the previous cycle — but the scale of AI-driven traffic raises the stakes well beyond any prior crawler dispute.

Three operational questions for wholesale leaders

• Channel classification. Will AI shopping assistants be treated as paid channel partners with negotiated economics, as neutral search engines with open access, or as unauthorized scrapers subject to technical blocking?
• Catalog integrity. How will product-attribute and listing standards hold up when AI agents re-rank inventory outside the host platform's user interface, stripping out the merchandising and promotional logic that drives attach rates?
• First-party data. What happens to distributor data assets once a third-party model trains on the catalog and resells the conclusion to the same buyers?

What to watch next

Distribution Brief will track whether the dispute produces a published access tier for AI agents, similar to the affiliate and API structures Amazon already publishes for sellers. The shape of that access — free, metered, or blocked — will determine whether AI becomes another demand channel feeding distributor catalogs, or a parallel commerce layer that captures the buyer's intent before the distributor ever sees the click.

The Digital Commerce 360 report signals the question has moved from theoretical to operational. Whichever platform sets the access terms first will own the unit economics of AI-driven B2B purchasing for the rest of the cycle.