There are certain lines of prompting that will result in heavily censored responses. They are quite obvious, and one could reasonably argue that it is a good idea to do so, but some of it seems rather arbitrary. It makes me wonder how much censoring is happening in the major AI tools that isn't obvious.

What are the best ways for an individual to get access to uncensored AI SAAS? I'd like to compare an uncensored AI to the majors.