AI companions are passive — they wait for prompts instead of proactively questioning kids to build reasoning

Summary

The core promise of the 'Kids News' concept is an AI companion that doesn't just deliver content, but actively asks a child questions, challenges their assumptions, and pushes them to explain their reasoning so screen time becomes active thinking rather than passive consumption. Yet today's AI agents are fundamentally reactive: users report that no matter how elaborate the persona or memory setup, the agent does nothing in the background and stays stuck in a 'person A, do this — person B, do that' mode, only responding when explicitly prompted. For a parent or teacher trying to run a 20-30 minute media-literacy discussion, this is fatal — the child is left to drive the conversation, which means they default to passive scrolling or silence rather than the Socratic back-and-forth that builds critical thinking. There is no product that reliably inverts this dynamic for kids: a companion that proactively initiates, remembers what the child said last week, surfaces an age-appropriate current-events prompt unprompted, and keeps the child arguing and reasoning. This is a severe, unsolved gap that directly blocks the entire 'active screen time / reasoning-building companion' use case, and it is the single biggest technical obstacle standing between families and the experience they are explicitly asking for.