If I were running a chatbot that was being criticized for problems with consent, I would make sure that it understands the word “no” without having to be blocked and reported.
I’d also make it so that it didn’t send unsolicited messages to everyone sharing articles about its ethics lapses as if they want to use it.
Even for an old-school chatbot this is just staggeringly incompetent and a gigantic flashing warning sign that the organization should not be trusted with sensitive data and high-stakes interactions.
