Over the holiday break, Elon Musk’s chatbot generated — then publicly posted on X, formerly Twitter — images of kids and other people who hadn’t consented in sexualized contexts.

It’s disgusting. And it’s happening on a platform that is still frequently used as a main channel for communication by Canadian governments of all levels. 

Someone prompted the X chatbot, Grok, for an apology; it generated a statement referring to the situation as “isolated cases.” Leaving aside the contradiction inherent in “cases,” plural, being “isolated,” the chatbot’s statement was the only substantive one the company behind it (Elon Musk’s xAI) issued. In what’s becoming a tired pattern in the AI world, the company did not take any responsibility for the actions of its product. 

This is far from the first time Grok has done something revolting, and it is guaranteed not to be the last. It’s a well-established pattern by now: Grok, after all, has been explicitly described and trained to be the “anti-woke” chatbot. It has

© National Observer