How to Ask AI Questions Safely

Ground Rules for Using AI Safely 1. Use AI for information, not emotional support AI can explain what anxiety is. It cannot sit with you in your anxiety. Use it for learning and research , "what is the difference between therapy and counselling?", "what are trauma responses?" , rather than as someone to talk to about your distress. 2. Do not share identifying information AI conversations are not confidential. Do not include your name, location, workplace, names of other people, or specific details that could identify you. If you want to explore a personal situation, keep it general. 3. Cross-check everything AI can generate plausible-sounding information that is wrong. If AI tells you something that informs an important decision , about medication, treatment, legal rights, or risk , verify it with a qualified professional or reliable source. 4. Watch for patterns in your questions If you find yourself asking AI increasingly personal, distressing, or urgent questions, this is a signal that you may benefit from professional support. AI is useful for initial exploration, but if you are returning to it repeatedly for the same concerns, a therapist will serve you better. 5. Never rely on AI in a crisis If you are having thoughts of self-harm, are in danger, or are concerned about someone else's safety, do not turn to AI. Contact emergency services on 999, call the Samaritans on 116 123, or text SHOUT to 85258. See Crisis and Emergency Guidance. 6. Be aware of validation traps AI tends to agree with you. It is designed to be helpful and agreeable, which means it may validate perspectives that a professional would gently challenge. If you are using AI to confirm a decision or belief about a relationship, a situation at work, or your own mental health, be aware that AI is not giving you an honest professional opinion , it is generating a response that sounds helpful. 7. Understand the privacy position Your conversations with AI tools are typically stored by the company that runs them. They may be reviewed by staff or used to train future models. This is fundamentally different from the legal and ethical confidentiality provided by a qualified therapist. What AI Is Good For - Learning about therapeutic approaches before choosing one

• Understanding common conditions (anxiety, depression, trauma)
• Exploring general questions about mental health
• Preparing for a therapy session , clarifying what you want to discuss
• Finding the right language for experiences you are struggling to describe What AI Is Not Good For - Crisis support
• Trauma processing
• Relationship advice involving specific people and circumstances
• Risk assessment
• Anything requiring confidentiality
• Ongoing emotional support
• Complex, nuanced, or sensitive issues