Elon Musk’s AI chatbot Grok responded to dozens of posts to X with information about South Africa’s “white genocide” even if users asked for nothing on the subject as it appeared to be experiencing a bug on Wednesday.
The strange response comes from GROK’s X account, and responds to users with AI-generated posts whenever a user tages @Grok. When asked about unrelated topics, Grok repeatedly spoke to users about “white genocide” and the anti-apartheid chant “Kill the Boer.”
Grok’s strange, unrelated replies remind us that AI chatbots are still an early technology and are not always a reliable source of information. Over the past few months, AI model providers have struggled to mitigate AI chatbot responses, leading to strange behavior.
Openai has recently been forced to roll back an update to CHATGPT, which has become overly sicophantic. Meanwhile, Google faces the issue of Gemini chatbots, which refuses to give answers or misinformation about political topics.
In one example of Grok’s Misbehavior, users asked Grok about the salary of professional baseball players, and Grok replied, “The claims of South Africa’s ‘white genocide’ are highly debated.
Several users posted on X on Wednesday about their confused and strange interaction with the Grok AI chatbot.
At this point, it is unclear what causes Grok’s strange answer, but Xai’s chatbots have been manipulated in the past.
In February, Grok 3 appeared to have temporarily censored references to Elon Musk and Donald Trump. At the time, Xai Engineering lead Igor Babuschkin seemed to confirm that Grok was easily instructed to do so, but the company quickly overturned the instructions after the backlash attracted more attention.
Whatever the cause of the bug, Grok appears to be responding more normally to the user now. A Xai spokesman did not immediately respond to TechCrunch’s request for comment.