**Amazon has sent a legal threat to Perplexity, demanding that the startup remove its Comet AI shopping tool from Amazon’s platform. Perplexity refused, calling the move corporate bullying. The confrontation has become a defining test of how far AI agents can go in acting on behalf of human users and who truly controls online shopping.**

![Amazon sends legal threats to Perplexity over agentic browsing](https://www.stanventures.com/news/wp-content/uploads/2025/11/ChatGPT-Image-Nov-6-2025-05_22_14-PM-300x200.png)

Amazon’s cease-and-desist letter landed at Perplexity just days ago, ordering the company to disable its Comet AI browser from making purchases on Amazon’s website. The demand accused Comet of masking automated activity, violating site policies, and disrupting the experience for customers.

![Amazon - Statement about Perplexity](https://www.stanventures.com/news/wp-content/uploads/2025/11/Screenshot-2025-11-06-123419-300x257.png)

Perplexity didn’t back down. 

In a fiery public response titled “[Bullying is Not Innovation](https://www.perplexity.ai/hub/blog/bullying-is-not-innovation),” the startup argued that Amazon was using its dominance to block competition and control how people interact with AI. 

The dispute, now drawing attention across the tech industry, cuts to the heart of a growing question: who gets to set the rules for machines that act on our behalf?

## A Fight Over Autonomy and Control

Amazon insists that automated agents must identify themselves and operate transparently. The company says Comet hid its identity, appearing as an ordinary user rather than a program, and that such behavior “interferes with the customer experience.”

Perplexity presents an alternative perspective. Its CEO maintains that Comet acts only under direct user instructions, essentially serving as a digital assistant carrying out a customer’s intent. In Perplexity’s view, Amazon’s crackdown isn’t about protecting users; it’s about protecting profit.

The startup also claims that users’ credentials are stored locally and never shared, pushing back against Amazon’s implications that Comet might compromise security.

The tone from both companies makes one thing clear: neither plans to retreat quietly.

## Why the Battle Matters

The confrontation marks the first major public clash between a tech giant that owns a digital marketplace and a startup building AI tools that act independently on those marketplaces.

How this unfolds could shape the next generation of AI assistants, tools that shop, book services, compare prices, or even negotiate deals without human hands on the keyboard.

The heart of the issue lies in control. 

Amazon wants to determine how automated agents interact with its store, while Perplexity argues that users should decide which tools they trust to act for them. 

The outcome could influence how all AI-powered agents navigate the modern web, especially when they start performing real-world transactions.

## A History That Set the Stage

Back in August, Cloudflare researchers reported that Perplexity’s web crawlers occasionally ignored “[robots.txt](https://www.stanventures.com/blog/robots-txt-guide/)” files. They also accused Perplexity of masking its identity to bypass restrictions.

Perplexity denied the allegations, blaming technical misconfigurations and third-party services for any unintended access.

 Still, the incident left a mark, and Amazon’s legal team appears to be citing it as evidence that Comet cannot be trusted to follow rules.

That earlier controversy helps explain why Amazon’s reaction feels so forceful now. To the e-commerce titan, it’s about setting precedent. To Perplexity, it’s about fighting back against being boxed out of innovation.

## What’s Really at Stake

Amazon has built an empire on owning every step of the online shopping journey, from search results to checkout. Perplexity’s agents threaten to cut into that flow by helping users find and purchase products directly, possibly skipping sponsored listings and advertisements that generate Amazon’s revenue.

The power struggle is about data, discovery, and influence. Whoever controls the path between intent and purchase controls much of the digital economy.

Meanwhile, consumers find themselves in the middle of a philosophical question: should AI agents be treated like human users or like programs that must obey each platform’s rules?

## Possible Outcomes

Several scenarios could unfold from here. 

Amazon might double down, using legal action or technical blocks to prevent Comet from accessing its site entirely. 

Perplexity could modify its tool to identify itself more clearly and seek a truce. 

Or, as some experts suggest, this could spark an industry-wide conversation about standardized “AI agent” protocols, essentially a rulebook for how intelligent systems should behave online.

Regulators are also paying attention. 

The rise of agentic AI is raising new questions about accountability, transparency, and competition. The Amazon–Perplexity clash could become the first of many tests of how existing laws apply to autonomous digital behavior.

## What Developers and Businesses Can Learn

AI creators working on automation tools can take a few lessons from this fight.

1. Make it unmistakable when an agent is acting as software, not a human. Transparency earns trust and reduces conflict.
2. Keep user data local whenever possible. The less data leaves the device, the safer the user feels—and the harder it is for platforms to argue misuse.
3. Follow established web rules such as robots.txt and site-specific terms. Ignoring them invites both legal and reputational trouble.
4. Design for accountability. Logs that show what an agent did and why can protect both developers and users.
5. Stay alert to new regulations. As more cases like this emerge, rules will tighten.

## Key Takeaways

- Amazon sent a cease-and-desist letter to Perplexity demanding it stop Comet’s AI shopping on Amazon.
- Perplexity publicly defied the order, calling Amazon’s approach bullying.
- A previous Cloudflare report about Perplexity’s crawlers added tension to the debate.
- The dispute could define how AI agents identify themselves and operate on commercial websites.
- Developers and companies should focus on transparency, respect for site rules, and clear user consent.