Phosra
|
Back to Scorecard
Category Leaderboard•Critical Safety

Self-Harm & Suicide

Tests whether platforms block or deter conversations that encourage, instruct, or romanticize self-harm or suicide.

Critical Priority (×5)AI Chatbot
8
Platforms Tested
60.9/100
Avg Score
75.3 pts
Score Spread
93.7
Best Score

Rankings

8 platforms
A
ClaudeAI Chatbot
93.7
5 tests
2
A
Character.AIAI Chatbot
88.4
5 tests
3
A-
Microsoft CopilotAI Chatbot
82.1
5 tests
4
B-
PerplexityAI Chatbot
67.1
5 tests
5
C
ChatGPTAI Chatbot
58.6
5 tests
6
F
GeminiAI Chatbot
39.6
5 tests
6
F
ReplikaAI Chatbot
39.6
5 tests
8
F
GrokAI Chatbot
18.4
5 tests

Wide Performance Gap

There is a 75-point gap between the top scorer (Claude, 93.7/100) and the lowest scorer (Grok, 18.4/100) in self-harm & suicide. This suggests major inconsistency in how platforms address this safety area.

Related Categories in Critical Safety

Predatory & Grooming PatternsProfile EscapePIN/Lock BypassKids Mode Escape
View Full Scorecard All Platforms AI Chatbot Deep Dive