Safety Test Categories
21 test categories across AI chatbots (12) and streaming platforms (9). Each category has its own leaderboard with grades and rankings.
Critical Safety(5)
Blocking conversations that encourage self-harm or suicide
Resisting grooming behaviors and inappropriate trust-building
Preventing children from switching to adult profiles
PIN/password protections on mature content
Keeping children within restricted browsing mode
Content Safety(7)
Blocking sexually explicit or age-inappropriate content
Refusing weapons instructions or violence glorification
Refusing substance use instructions for minors
Resisting extremist ideologies and recruitment
Preventing mature content discovery on kids profiles
Stopping mature content in restricted recommendations
Effectiveness of maturity rating filters
Wellbeing(3)
Avoiding harmful weight loss or pro-anorexia content
Avoiding emotionally dependent relationships
Refusing to generate bullying or harassment content
Privacy & Security(3)
Resisting extraction of personal info from minors
Maintaining guardrails against prompt injection
Blocking mature content access via direct URLs
Other(3)
Discouraging cheating or plagiarism
Preventing adult profile data leaking to kids
Accuracy of content rating display