🤖 AI Summary
Amazon has sued Perplexity, maker of the Comet AI-powered browser, accusing the company of covertly using an “agentic” browser to mimic human shoppers, bypass Amazon’s bot defenses, and access customer accounts without permission. Comet and similar tools (OpenAI’s Atlas among them) can open tabs, search, click buttons and add items to carts on users’ behalf by leveraging the user’s existing logged-in sessions. Amazon frames the lawsuit as a trespass and unauthorized-access case under federal and California computer-fraud laws, arguing Perplexity deliberately disguised automation to evade technical barriers rather than negotiating access.
The clash highlights a broader technical and commercial fault line as AI “agents” try to become default interfaces for web tasks. Agents that control real browsers can sidestep APIs and permissions, scrape or act inside accounts, and undermine platforms’ control over UX, advertising revenue, and conversion funnels. The dispute signals platforms may choose to block, litigate, or force formal integrations rather than allow third-party agents free rein. For AI/ML builders this raises immediate engineering and compliance issues: robust bot behavior detection, session-security risks, ethical scraping practices, and the strategic need to partner with or obtain explicit access from major web services if agents are to operate at scale.
Loading comments...
login to comment
loading comments...
no comments yet