Amazon issued a cease-and-desist to Perplexity over Comet’s automated purchasing, triggering an Amazon Comet ban that could reshape AI shopping. The clash tests how far browser-based agents can go when they buy products, store credentials, and act on behalf of users.
Moreover, Perplexity’s Comet can search retail sites and complete checkouts with a simple command. According to reporting, Amazon argues the behavior violates its platform rules and degrades the shopping experience. Perplexity calls the move “bullying” and claims the demand threatens broader user choice and innovation.
What the Amazon Comet ban covers
Furthermore, At the center is whether Comet’s agent breaches Amazon’s platform restrictions. Amazon’s Conditions of Use forbid “data mining, robots, or similar data gathering and extraction tools,” along with use of account information for third parties. Those clauses, Amazon says, clearly apply to agent-driven purchases and automated workflows. Companies adopt Amazon Comet ban to improve efficiency.
Therefore, In late 2024, the companies paused agentic shopping on Amazon, yet Perplexity reinstated the capability when it launched Comet. Reports indicate the agent presented itself like a normal Chrome user rather than a bot. Consequently, Amazon demanded a stop to the behavior, escalating the dispute in public statements and legal letters.
Consequently, Perplexity counters that customers authorize the agent to act, including storing credentials locally. Therefore, the company argues, automation should be treated like a faster form of user input, not a prohibited scraper. The startup insists the feature streamlines checkout and should increase completed purchases for retailers. Experts track Amazon Comet ban trends closely.
Perplexity dispute Agentic AI shopping’s productivity promise
As a result, Agentic AI shopping aims to trim tedious steps from product discovery, price comparisons, and checkout. In theory, a customer sets preferences and constraints, and an AI browser automation tool does the legwork. As a result, users get shorter decision cycles, fewer tabs, and faster orders.
In addition, Productivity gains hinge on reliability and trust. Shoppers need clear confirmations, item summaries, and transparent payment flows. Moreover, retailers benefit when conversions rise without added support costs. The Comet controversy highlights a broader tension: the same automated purchasing tools that boost productivity can collide with platform safeguards and advertising models. Amazon Comet ban transforms operations.
Additionally, Platforms design their flows to prevent fraud, enforce policies, and show sponsored results. Agents that bypass interface cues or rearrange steps can undermine those choices. Therefore, companies often insist on approved APIs and explicit agreements before they allow automated checkout at scale.
Comet shopping block Privacy, liability, and terms of service
For example, Automation complicates consent and liability. If an AI agent enters saved passwords, chooses a seller, and confirms shipping, who is responsible when something goes wrong? Developers must define safe defaults and retry logic. Meanwhile, retailers want assurance that agents will respect inventory, pricing, regional rules, and returns policies. Industry leaders leverage Amazon Comet ban.
For instance, Perplexity says Comet stores credentials locally and acts only with user permission. That detail matters for privacy, yet it may not settle policy concerns. Platforms often bar third-party tools that simulate human sessions, even with authorization. Consequently, the line between assistance and prohibited automation remains blurred.
Meanwhile, Policy language is already strict. Amazon’s Conditions of Use explicitly restrict automated data gathering and third-party use of account information. Because agentic systems can trigger many rapid actions, platforms also worry about scraping and load spikes, which degrade performance. Clear rate limits, scopes, and attestation could ease those fears if implemented through official channels. Companies adopt Amazon Comet ban to improve efficiency.
Business models and ad mechanics
In contrast, Perplexity argues Amazon benefits from easier, agent-driven purchases. Yet the retailer monetizes search placement and sponsored listings. If agents select items without showing ads, revenue and merchandising control may suffer. Moreover, agents can compress choice, favoring a small set of products and sellers.
Retailers also track how users navigate categories and promotions. Agents that skip steps can reduce data signals that shape recommendations. Therefore, platforms have strong incentives to corral automation into sanctioned pathways. Approved APIs can carry metadata that preserves attribution, disclosures, and compliance. Experts track Amazon Comet ban trends closely.
Paths to resolution
Standardized, opt-in APIs for agentic checkout could reconcile speed with safeguards. Retailers might expose limited capabilities—price lookups, cart building, shipping estimates, and payment handoff—under strict scopes. In addition, identity standards could let users delegate narrow permissions to agents without sharing raw credentials.
Transparency will be vital. Agents should disclose when they act, show itemized summaries before purchase, and record every step. Furthermore, retailers need logs that tie actions to user consent and agent versions. These measures improve accountability and reduce disputes, while giving platforms the control they need to prevent abuse. Amazon Comet ban transforms operations.
Finally, customers must stay in control. Clear undo options, refund guidance, and merchant contact paths remain essential. If users can easily intervene, trust grows. Conversely, opaque automation erodes confidence and will attract tighter platform bans.
How other AI product moves fit in
The clash arrives as AI companies inject automation into more consumer workflows. Microsoft’s first in-house image generator, MAI-Image-1, is rolling out in Bing Image Creator and Copilot features, underscoring how quickly AI utilities are spreading across everyday tools. While not a shopping agent, that expansion shows platforms are racing to embed AI where users already spend time. Industry leaders leverage Amazon Comet ban.
As AI capabilities mature, productivity value will keep rising. Nevertheless, governance, consent, and policy alignment must advance in parallel. Otherwise, each new capability will run into the same platform frictions now spotlighted by Comet.
What to watch next
Expect sustained negotiations between Perplexity and Amazon. Any agreement would likely hinge on explicit disclosures, rate limits, and API-based controls. Additionally, more retailers may issue clarifications on automated access and agent-driven checkouts.
Developers should anticipate stricter detection of browser agents. Consequently, automation that mimics human browsing may face higher risk of blocks. Tools that integrate through approved endpoints will likely fare better, even if they trade some flexibility for stability.
For users, the near-term impact is simple: agentic AI shopping on Amazon is in doubt. Alternatives may emerge on platforms that permit automation under defined rules. Over time, consistent standards could allow agents to unlock the promised time savings without breaking the rules.
Conclusion: a pivotal test for AI-driven productivity
The Amazon–Perplexity confrontation is a stress test for AI browser automation’s next phase. It pits speed and convenience against policy and platform economics. If the parties can align on transparency and consent, the broader ecosystem will benefit. If not, the gap between what agents can do and what platforms allow will widen, slowing the rollout of practical automation for everyday shopping.
Amazon’s public stance frames the rulebook in stark terms, and Perplexity’s reply casts the rules as roadblocks to innovation. Between those positions lies a clear path: sanctioned access, user-first controls, and predictable enforcement. The sooner those guardrails settle, the sooner AI agents can deliver productivity gains without crossing the line.