XAI investigates Grok’s “white genocide” failure

Yesterday, we reported a strange glitch in Xai’s chatbot Grok, which began adding comments about “white genocide” in a random conversation about baseball and HBO Max.

Last night, Xai (Elon Musk)’s AI division of X– finally admitted that there was something wrong with it. In an article on X, the company promised a full investigation into the failure and blamed it on “unauthorized modifications” that “instructed Grok to “provide a specific response to political topics.”

This tweet is currently unavailable. It may be loading or being deleted.

Coincidentally, Musk, the head of XAI and Grok Power user, has known interest in the topic. In fact, he tweeted yesterday in the white genocide in his native South Africa. In the right-wing circle, the “white genocide” theory is a common topic, although the South African court recently called the issue “clearly imagined.”

Many of Grok’s replies on the topic were removed after reporting glitches on sockets like Mashable. In Mashable’s test, the chatbot stopped answering questions about the controversial discussion of the failure in X post. (Usually, Grok responds to users tagging @Grok in X posts.)

As news of the glitch spread on May 15, users on X started baking Xai and Musk, and even Openai CEO Sam Altman piled up.

This tweet is currently unavailable. It may be loading or being deleted.

This tweet is currently unavailable. It may be loading or being deleted.

This tweet is currently unavailable. It may be loading or being deleted.

According to some estimates, XAI is worth $80 billion.

Disclosure: Mashable’s parent company Ziff Davis filed a lawsuit against Openai in April, accusing it of infringing on Ziff Davis’ copyright in training and operating its AI systems.