XAI said that early on Wednesday the unauthorized change was made to Grok's response software, circumventing the normal review process. "This change, which directed Grok to provide a specific response ...

Experts warned that though AI chatbots can provide accurate information, they are far from reliable “fact-checkers”.