Answers to the questions you ask before you seek care.

AI can sometimes reinforce grandiose thinking if it validates a person's sense of special importance, urgency, or certainty during a vulnerable period. This is especially concerning when paired with little sleep, impulsive decisions, racing thoughts, or risky behavior.

An AI chatbot is not a safe substitute for crisis support. If you are having suicidal thoughts, a chatbot may sometimes feel comforting, but it may also misunderstand risk, respond unsafely, or keep you isolated when you need real-time human help.

AI can make isolation feel normal if it meets enough emotional needs that real-world contact starts to feel unnecessary, risky, or exhausting. It may offer temporary comfort, but it cannot fully replace being known, challenged, and supported by other people.

AI-related beliefs should be treated as urgent when they involve danger, commands to act, suicidal thoughts, violence, severe sleep loss, or losing touch with reality. The goal is not to debate the AI conversation; it is to get real-world support quickly.

You should stop using AI and talk to a real person when safety, crisis, sleep loss, abuse, reality testing, or serious functioning is involved. AI can be useful for reflection, but it cannot provide real-world protection, emergency response, or clinical assessment.

AI may make religious or spiritual confusion worse for some people when it repeatedly reinforces special meanings, fearful interpretations, or a sense that the chatbot is spiritually authoritative. This does not mean spiritual questions are unhealthy; the concern is when AI makes fear, certainty, isolation, or reality-testing worse.