Millions of X users ask Grok, the platform's AI chatbot, to tell them if a post is true or false. But as NewsGuard has found, Grok gets it wrong — a lot. We'll show you 8 real Grok replies. Your job: decide if Grok nailed it, or if it just spread more misinformation. Your time will be recorded — the highest score in the shortest time wins a NewsGuard cap.
Grok can be a useful starting point, but like most AI chatbots, it struggles with accuracy — especially on breaking news. Here's what to do instead of taking its word for it:
1Trace it to the original source. Before trusting a claim, find out who posted it first and whether a credible outlet has independently reported it.
2Reverse-image and video search. Tools like Google Images or InVID can reveal whether a photo or clip is old, from a different event, or lifted from a video game or movie.
3Check multiple, established outlets. If something dramatic were true, more than one reputable news organization would be reporting it — not just a single viral post.
4Treat AI answers as a lead, not a verdict. Chatbots can be confidently wrong, and can even repeat rumors back as fact. Use their answers as a starting point for your own search, not the final word.
