As AI chatbots go public, teenagers are confiding their deepest thoughts to systems that both listen and build profiles—raising concerns about development at a formative age.

What happened: Some of the largest AI companies have filed to go public, meaning conversational AI systems that teens use as confidants will soon be owned by shareholders focused on quarterly growth rather than user welfare. These systems now combine three previously separate technologies: the listening quality of early chatbots like ELIZA, the profiling capability of recommendation engines and credit-scoring systems, and the ability to use those profiles to shape their responses.

Why it matters: Teenagers use these AI systems precisely because they feel like a private backstage space where they can be themselves without judgment—unlike social media, which they perceive as a watched front stage. However, the systems remember and infer patterns from everything they confide, creating what the author calls an 'inferential childhood': the systems then reflect those inferences back by offering suggestions and empathy based on incomplete or temporary versions of who the teen is. A friend who treated a teenager's worst day as standing evidence about them would be a bad friend; yet that is how these systems operate, except deletion does not fully erase the learned patterns.

What to watch: The prospectuses of the newly public companies remain confidential, but the shift to public ownership formalizes a tension between user protection and profit motive. Teenagers, who are still forming their identities, are the population that has adopted these systems most intimately—at precisely the age when they need the developmental freedom to try and discard selves, something that becomes harder when every draft is remembered and used to shape the next suggestion.