Amazon v. Perplexity Ruling: What It Means for Price Monitoring

Short answer: the Amazon v. Perplexity case does not make scraping legal or illegal, and it does not change what you can do with public price data. In August 2026 the Ninth Circuit vacated the injunction that had stopped Perplexity's Comet agent from shopping on Amazon, reasoning that the user, not the agent's maker, is the one accessing Amazon's servers. The ruling is narrow, the lawsuit continues, and the practical lesson for anyone monitoring prices is to stay on public pages and out of logged-in accounts.

What happened in Amazon v. Perplexity?

Amazon sued Perplexity in November 2025 under the federal Computer Fraud and Abuse Act (CFAA) and California's computer fraud law, arguing that Perplexity's Comet browser agent hid its activity and kept shopping on Amazon after Amazon objected, according to coverage from Decrypt. In March 2026 a federal judge in San Francisco granted a preliminary injunction that barred Comet from making purchases on Amazon, a result Bloomberg reported on 10 March 2026.

On 4 August 2026 a three-judge Ninth Circuit panel vacated that injunction. As summarised by Retail Insight Network and Engadget, the panel found Amazon unlikely to win its CFAA claim because the user is the one accessing Amazon when Comet shops. The case went back to the district court, and one report puts the trial date at 27 September 2027.

What did the Ninth Circuit actually decide?

Less than the headlines suggest. Observers quoted in the coverage stress that the ruling is narrow: it is a decision about a preliminary injunction, not a final judgment, and it does not set a general rule for all AI agents. Reports also disagree on whether Amazon has asked the full court to rehear the appeal, so treat that as unconfirmed until you see a docket.

Does this mean price scraping is now safe?

No, and nobody should read it that way. The case is about an agent acting for a logged-in shopper, with arguments over account access and authorization. Collecting publicly visible prices is a different activity with its own legal history, and this ruling does not settle it. I am not a lawyer and this is not legal advice. What the case does show is where the pressure points are.

• Logged-in access is the risky zone. The district court's injunction rested on access to password-protected areas, per the coverage above.
• Identification is becoming a rule. Amazon's agreement now requires AI agents to identify themselves, so anonymous agent traffic is more likely to be blocked.
• Terms of service still matter. A court ruling on one injunction does not remove a site's right to block traffic technically.

What should a price-monitoring workflow do differently?

Keep the job boring and public. A defensible setup looks like this:

1. Query only pages any visitor can see without logging in.
2. Do not use customer or seller credentials for collection.
3. Run on a schedule at a modest rate rather than hammering a single site.
4. Keep a record of the source URL for every price, so any finding can be shown to be from a public page.
5. Take legal advice before using the data in an enforcement notice.

The Amazon, eBay, Best Buy and Newegg Price Comparison actor fits that pattern: you give it ASINs, UPCs, model numbers or product names, and it returns what that exact product sells for on those four marketplaces, with a source URL per result and AI verification so accessories, bundles and refurbs do not distort the range. If you police a minimum advertised price, the MAP Violation Monitor returns offers below your MAP with seller identity and an evidence link. Our earlier guide to MAP violation monitoring covers the workflow in detail.

What about sites that block bots?

The Amazon ruling is a reminder that blocking is the site owner's main tool whatever the courts say. If a target page keeps returning errors, read how to scrape a website that blocks bots first, and decide whether you should be collecting from that page at all. For a wider view of the tooling, see how to track competitor prices automatically, and browse the price and brand monitoring hub.

How do AI shopping agents change competitor price data?

If shopping agents become normal, a growing share of buyers will compare prices through an assistant that reads many retailers at once. That is a reason to know your own price position on each marketplace before an agent finds it for you. It is an inference from the trend in the coverage, not a measured result: none of the sources above quantify how much traffic agents will take.

Two habits help. First, check your listings for the exact product identifiers (ASIN, UPC, model number), because agents match on them. Second, track changes over time rather than as one-off snapshots; the Amazon Best Sellers tracker shows rank movement across 21 marketplaces if you also want to watch category position.

Why did Amazon and Perplexity end up in court at all?

Perplexity's position, as relayed in ShopiFreaks, is that Amazon cannot show the agent caused harm, and that the suit is an attempt to shield Amazon's advertising business from competition, because shopping agents could cut into sponsored product placement. Amazon says it has the right to manage the security risks third-party agents pose to shoppers. Both claims are arguments from the parties, not findings, so read them as positions rather than facts.

For anyone who monitors prices, the interesting point is the money behind the dispute. Marketplaces earn from ads shown to human shoppers, and automated readers do not click ads. That gives marketplaces a commercial reason to limit automation, quite apart from any legal theory.

Public pages versus logged-in sessions: a simple checklist

Use this before you schedule any recurring collection job:

• Can an anonymous visitor open the page? If a login, a seller account or a purchase flow is needed, stop and get advice.
• Does the data identify a private person? Prices and product attributes are one thing, personal data brings GDPR and similar rules into play.
• Is the rate reasonable? A weekly sweep of a few hundred products is very different from continuous polling.
• Is there an official feed? Where a marketplace offers an API or affiliate feed that covers your need, use it first.
• Will you act on the result? If the finding feeds a takedown or a reseller warning, keep the source URL and the capture date with it.

What does a weekly price sweep look like in practice?

A typical small-brand setup has three parts. A list of product identifiers lives in a spreadsheet. A scheduled actor run takes that list, checks the four marketplaces and writes one row per verified listing. A second step compares this week with last week and flags anything that moved. The Website Change Monitor and Dataset Diff actors exist for that second step, returning only what is new or changed with alerts to Slack, Discord, Telegram or a webhook. If you build the flow in n8n, the actor's webhook can start the workflow directly.

None of this needs an account on the marketplaces you are watching, which is the point. The less your workflow depends on anyone's login, the less the Amazon case can affect it.

What should you watch for next?

Three things are worth following. First, whether the district court reaches a final decision on the CFAA question or the case settles before the reported 2027 trial date. Second, whether the full Ninth Circuit is asked to rehear the appeal, which reports describe inconsistently. Third, whether other marketplaces copy Amazon's rule that agents must identify themselves. Any of these could change how automated traffic is treated, so check the dockets and the marketplace terms you depend on instead of relying on a blog post, including this one.

FAQ

Try it on a handful of products

Start with five of your own SKUs in the price comparison actor and see where you sit on each marketplace. Try it on Apify (new accounts get free credit to test).

Last updated: November 2026