AI agents with friendly faces and adorable bodies can distract you from serious risks

The authors argue that tech companies are using cute designs and friendly avatars, such as Meta's Muse and OpenAI's Dots, to distract users from serious privacy risks and extensive data collection. By evoking emotional attachment and trust, these disarming features encourage users to grant agents broad permissions to access messages, calendars, and accounts without skepticism. The authors suggest that companies should provide clear, plain-language disclosures about what these agents can access and do. Open article →