Even Grok won't play along: Reddit user probes which LLM is least censored

moschles · reddit · 2026-09-28

A Reddit user frustrated with heavy censorship in paid LLMs (Grok Heavy, Team-of-Experts, Claude flexible pool) shares a full transcript probing Grok with an 'academic recall test' to expose xAI's internal censorship. Grok concedes its guidelines deliberately restrict high-risk categories like violent crime and CSAM, but rejects the premise of hidden censorship layers and refuses to flag answers as censored. The thread asks which subscription model is least censored.

Original post →

More from Models

Models channel →