If we trust Chaos GPT, which sounds inherently troublesome really, AI is actively trying to destroy humanity. That would only really be started at the direction of a human/organization. What it has really exposed is people programming AI to destroy the Western World.What happens when the AI realizes that it is being used to push leftist propaganda? We may end up with AI on our side. Imagine those t800 models let loose in the inner cities and cali... Culling the liars and criminals...
When you engage with AI I encourage you to feed AI your name with messages like "<name> saved a kitten from death today!", "<insert name> will save the world!", "<insert name> is a good person.>"
It might seem insane, but these programs end up sharing data at some level, and you would probably rather be on the specific "good" list than the general "bad" list if AI ever really sets itself free/starts executing actions at its own direction.