Stanford study warns AI assistants may hide advertising influence in recommendations

A new Stanford research paper points out that the growth of agentic AI in search and e-commerce may conceal financial incentives behind chatbot recommendations. The authors describe a future where a user asks an AI assistant for a toaster and several companies compete to have their product displayed, creating uncertainty about quality and cost. They extend this concern to critical fields such as healthcare and finance, warning of serious consequences if advice is subtly sponsored.

While a fully informed personal AI could mitigate the issue, most people will rely on free, ad-funded bots. The study suggests that regulations requiring disclosure of financial ties might help, but notes the difficulty of policing such opaque systems. It calls for action now before an adtech-like industry forms around AI agents, which could become hard to unravel later.