So far no drug has been approved to treat the 'core symptoms' of autism spectrum disorder. Those are related to social isolation and lack of communication. This study suggests to me (I didn't read the technical article) that if such a medication family were found it could also benefit people with bipolar or schizophrenia.

My guess is that this hypothetical autism medication would be able to treat the so-called 'negative' symptoms of schizophrenia and related symptoms in bipolar.