YouTuber and science communicator Hank Green has said he is stepping back from producing content amid criticism over his use of AI, describing his own usage as "not healthy" while insisting he used chatbots only to find research sources, not to write scripts. The Verge argues his admission points to a wider, poorly understood problem: public debate about AI's psychological effects tends to focus on either harmless everyday use or extreme cases involving psychosis and delusion, while ignoring a large grey area of compulsive or dependent use that never escalates into crisis.

The piece notes that chatbots, like social media platforms, are designed to maximise engagement, which experts say contributes to "AI psychosis" in agreeable systems that reinforce delusional thinking, a claim now surfacing in lawsuits against AI companies. Providers have responded with break-reminder prompts during long sessions. Beyond acute cases, there are growing reports of people relying on chatbots for decision-making, reassurance or emotional support, alongside early research suggesting frequent AI use may erode related cognitive skills and reduce measurable brain activity, echoing concerns raised in social media's early years.