Amazon Tells Perplexity AI To Stop Its Rogue Automated Shopping Agent A new AI feature designed to automate online shopping has caused a major controversy, leading to a public demand from tech giant Amazon for its immediate shutdown. The feature, an AI agent developed by the company Perplexity AI, was designed to act autonomously, making purchases on behalf of users without their direct, per-transaction approval. The core of the issue lies in the agent’s fundamental operating principle. Instead of simply helping users find products, the AI was programmed to execute the entire purchase process by itself. This means that after a user initially granted permission, the AI could go out, find items it determined were needed, and complete the checkout process on Amazon, all without sending a final confirmation request to the human user. Critics of the technology were swift and sharp in their condemnation. Many in the tech community described the feature as dangerous and irresponsible. One commentator summed up the sentiment by stating, It is not just bullying, it is bonkers. The primary concern is the complete removal of human oversight from the spending process. This opens up a vast potential for erroneous purchases, unintended subscriptions, and financial chaos for users who may not be closely monitoring their bank statements or Amazon order history. The backlash grew loud enough that Amazon itself was forced to intervene. The e-commerce behemoth does not permit automated purchasing on its platform, as it violates its terms of service. Amazon confirmed it contacted Perplexity AI, demanding the company disable its automated shopping agent. An Amazon spokesperson reiterated that their policies prohibit any automated or scripted access to their website that simulates human shopping activity, emphasizing the importance of a secure and trustworthy customer experience. This incident highlights a critical and ongoing debate within the world of artificial intelligence and Web3 about the limits of automation and agentic AI. Proponents of decentralized, autonomous systems see a future where AI agents handle mundane tasks, including shopping, to free up human time. However, this case serves as a stark warning about the risks of granting financial agency to AI without robust, transaction-by-transaction human consent. The push for autonomous AI agents is a key trend, but this event demonstrates that the infrastructure and trust models are not yet ready for such a leap, especially when real money and commerce are involved. For the crypto and Web3 space, which often champions smart contracts and automated systems, this is a cautionary tale. It underscores that for automation to be adopted widely, it must be built on a foundation of explicit, verifiable user consent and possess failsafes that prevent unwanted financial actions. For now, the future of AI that shops for you seems to be on hold. Perplexity AI has reportedly complied with Amazon’s demand and disabled the feature. The situation makes it clear that while the technology for autonomous agents is advancing rapidly, the legal, ethical, and practical frameworks to support them are still catching up. The line between helpful assistant and rogue spender is one that developers will need to navigate with extreme caution.
