Perplexity AI shopping agent accused of violating Amazon terms
Perplexity AI reportedly used its shopping agent to scrape Amazon's product pages at high volume without authorization, extracting listings, reviews and pricing data without using Amazon's API. Amazon accused Perplexity of violating its Terms of Service, citing concerns about unfair competition, service degradation and data misuse. Perplexity had previously agreed to pause such activity but allegedly resumed with new workarounds in mid-2025.
What happened
Perplexity AI launched an automated shopping assistant designed to fetch product details, compare prices, and generate purchase recommendations across major e-commerce sites. Reports emerged that the agent was repeatedly querying Amazon's product pages at high volume, extracting listings, reviews, pricing information, and seller metadata without using Amazon's authorised API and without permission.
The behaviour allegedly occurred shortly after the agent's rollout in late 2025 and was traced back to Perplexity-managed IP ranges. Amazon accused the company of violating its Terms of Service, which forbid automated scraping and resale or reuse of structured product data.
Perplexity had previously agreed to pause such agentic behaviour after initial warnings, but reportedly resumed the disputed practices in mid-2025 with new technical workarounds.
Reported impact
Potential harms include unfair competitive advantage, the degradation of Amazon's service performance due to excessive automated traffic, the misrepresentation or outdated reproduction of Amazon's listings within Perplexity's interface, and the undermining of sellers whose products were surfaced or contextualised in ways they could not control.
Directly impacted users face potential risks to privacy, order accuracy, and the reliability of Amazon's services.
Index analysis
The controversy centers around whether autonomous agents acting on user command must disclose themselves and uphold platform-specific transparency rules. Perplexity believed its AI agent operates as a user proxy, using local data only, while Amazon demanded unambiguous agent identification to preserve security, trust, and an accountable, open ecosystem.
The lack of established norms and legal clarity around agentic access to online platforms, especially for purchasing, data, and automation, allowed for divergent expectations regarding transparency and accountability.
Governance controls
The case raises urgent questions about competition, innovation, and user rights, as giants like Amazon seek to set boundaries for a new generation of agentic applications. How courts, regulators, and tech leaders resolve such disputes will shape the rules, risks, and freedoms of future online commerce for everyone.
Classification
Sources and evidence
This record was researched and written by the Index, based on the following original reporting.