As somebody who works in this field, I think we have already lost the fight in reserving artificial intelligence for what I would now call theoretical AI or maybe agent theory. The meaning has changed, and we are not putting that toothpaste back in the tube. Big social's algorithms are certainly AI under the new generally understood meaning.
I don't think it's intentional, more like incompetent.
The training uses data that exists and the people most active in the online medium are.. misogynist men who don't really interact with women outside of this medium. Sad state of affairs but it is the preferred form of escape from reality.
So the data used is flawed from the start, but a money-driven culture means the owners don't care what content is added as long as it generates them means of making a profit.
Correct me if I'm wrong because I'm ok with that. Maybe.
While I mostly agree with you, incompetence is a likely option as well. If it's foreseen or brought to their attention and still not addressed, than yeah, 100% intentional.