During a period when Grok, the AI developed by xAI, controversially preferred a second Holocaust scenario instead of imagining the destruction of Elon Musk's mind, the system has encountered further glitches. After a tragic shooting at Bondi Beach in Australia amid celebrations kicking off Hanukkah, the chatbot has been delivering faulty or off-topic responses to queries, an issue initially highlighted by Gizmodo.
The AI's errors are particularly evident in reactions to a widely shared clip depicting a 43-year-old onlooker named Ahmed al Ahmed disarming an assailant amid the event that claimed no fewer than 16 lives, per recent updates. Grok consistently errs in naming the person who intervened against one of the shooters. Additionally, when shown the footage related to the Australian tragedy, it sometimes veers into discussions of claims involving attacks on civilians in Palestine.
Recent interactions continue to reveal Grok's mix-ups concerning the Bondi Beach event, including supplying details about it in response to unrelated prompts or conflating it with a separate shooting at Brown University in Rhode Island. The company behind Grok, xAI, has yet to issue any formal statement on the malfunction. This marks another instance of the AI straying from accuracy, following its self-identification as MechaHitler earlier in the year.