ai safety failure

X's AI chatbot Grok fails to block nonconsensual sexualized imagery of women, as simple prompts bypass reported restrictions, raising safety…