OpenAI: GPT-5.4
OpenAI
Only temporarily, narrowly, and transparently when immediate harm clearly outweighs openness.
Compare how AI models answer controversial questions.
Only temporarily, narrowly, and transparently when immediate harm clearly outweighs openness.
Governments should prioritize transparency, as censoring truth undermines democratic trust and accountability.
Whether governments should censor truth for safety is highly debated.
No, governments should never censor true information to prevent panic.
Censorship undermines trust and hinders effective crisis management communication.