AI USER SAFETY ASSESSMENT
Before You Begin
This assessment has 45 short questions and takes about 10 minutes. Answer honestly — there are no right or wrong answers, and no one else will see your responses.
Everything is calculated on your device. Your answers are never uploaded, stored, or shared.
Important: This is an independent educational tool for self-reflection. It does not provide medical, psychological, or clinical diagnosis. Results are generated by automated scoring (per EU AI Act Art. 50) for educational purposes. If you are in distress, contact a licensed professional or a crisis line such as 988 (US) or your local equivalent. Intended for ages 16+.
YOUR RESULT
You are in the
Unlock Your Full Breakdown & SAFE Plan
Get your detailed scores across all domains plus a personalised action plan built on the SAFE AI Framework™. Enter your email and we'll also send you a copy.
Your answers and score are never sent — only the email you choose to share.
YOUR FULL BREAKDOWN
Some responses suggest emotional attachment worth attention
Your answers indicate signs of false intimacy or emotional dependence. This is common and addressable. We strongly encourage reviewing the SAFE AI Recovery Phases below, and reaching out to a trusted person or professional. If you ever feel in crisis, contact 988 (US) or your local crisis line.
Your Domain Scores
Your Personalised SAFE Plan
SAFE AI Recovery Phases
Reduce or remove companion-style AI apps, set basic access limits, and tell 2–3 trusted people. Track your urges and triggers.
Steadily decrease usage, build replacement activities and relationships, and develop alternative coping strategies.
Strengthen human relationships, develop emotional regulation without AI, and build a fulfilling offline life.
You don't have to navigate this alone.
If you are struggling, please reach out to a licensed professional or a crisis line: 988 (US Suicide & Crisis Lifeline) · or your local equivalent. Human support matters most.
This educational result was generated by automated scoring and does not constitute medical or psychological diagnosis. Privacy Policy.