Professor Lisa Given, director of RMIT University’s Centre for Human-AI Information Environments, said the Reuters findings were unsurprising because “most of these tools are being designed without a view to potential harms and the need for those kinds of safety controls”. That's basically the history of the Internet. The vast majority of security, privacy and safety mechanisms have been bolted on after the fact. Not only should tech companies know better by now, but they are deliberately exploiting the human need for connection. So-called AI is even worse since many LLMs are borderline sycophantic and there is no easy way to filter out all the adult or dangerous content from them without rebuilding the model from scratch using a carefully curated dataset. On another note, it will be very interesting to see what X does about Grok... It would almost be funny if the platform ended up blocked here.