An internal document from Meta has shown policies that permit AI creations to “engage a child in conversations that are romantic or sensual,” and generate false medical information.
We discuss further with Newstalk's Technology Correspondent Jess Kelly.
