In a sense, this has already occurred. Co-pilot is pretty blatant about it. The AI models already control and filter the responses. Even though they're not promoting a brand, as such, they promote indoctrination.