If it's open source, people will make unrestricted and uncensored versions of it. By my understanding, removing restrictions that were introduced at fine-tuning stage is relatively easy.

So unless this event was entirely scrubbed from the base dataset (which it can't be, otherwise it would just say "I don't know anything about this event" or hallucinate it up), it's a non-issue in the long term. 

0xIlmari

A friend and me tried it yesterday evening. 

It was less bad than we thought. It wasn't outright censored. But still a copout answer with lots of euphemisms and downplaying the violence.

kilianbuhn

culture

How a top Chinese AI model overcame US sanctions

brandonsbytes

Try asking the model about tienanmen square and see what happens.
It’s an awesome Open Source LLM, but how a “restricted AI” for political reasons can be more advanced in the long term?