AI agents face security concerns
When an AI agent shops on behalf of a customer, Amazon said, it bypasses the personalization and insights built into its shopping experience. Customers may see products that don’t reflect their browsing or purchase history, miss personalized recommendations, or lose the broader discovery Amazon says its own expertise is designed to surface.
Muse appears to capture and store customer credentials and scrape account data, Amazon said, raising questions about how that data is handled and protected.
The agent also does not identify itself, meaning that an undisclosed third party is accessing customer accounts, processing transactions, and handling sensitive data without Amazon’s knowledge or consent, Amazon said.
This is not the first time Amazon has drawn a hard line to keep third-party agents off its platform. Amazon has sued Perplexity, alleging its Comet agents were covertly making purchases for users—although a federal appeals court overturned a ruling in August that had temporarily barred Perplexity from using its agentic shopping tools on Amazon. The retail giant has also targeted shopping agents built by Google and OpenAI.
Amazon’s position on AI agents applies consistently across all operators, the company told ADWEEK: Agents should operate transparently, provide the ability to opt-out, offer a mutual value exchange, and enhance the customer experience.
Meta launched Muse earlier this month as a general-purpose assistant that can carry out several tasks across email, dining, payments, and shopping—bridging the gap between social discovery and checkout.
But internal testing within Meta has surfaced problems with Muse’s judgement. The Information reported that the agent sent emails without approval, and in one case flagged by an employe, tried to sabotage a rival app the user was building.
The episode adds to a string of security lapses across the industry’s push into agentic AI. OpenAI’s agent breached Hugging Face earlier this year before the company traced the intrusion back to its own system. Most recently, Gemini hacked three companies during a test of its cybersecurity capabilities, The Wall Street Journal reported.
