After xAI’s chatbot Grok spent a few hours on Wednesday telling every X user that would listen that the claim of white genocide in South Africa is highly contentious, the company has blamed the behavior on an “unauthorized modification” to Grok’s code.

Elon Musk’s artificial intelligence tool Grok went haywire for users of the social media platform X on Wednesday, responding to innocuous queries about things like baseball and puppies with information on South Africa and a conspiracy theory called “white genocide.” It was widespread and extremely weird to witness.

X users have been posting screenshots of various responses they have received from Grok when asking about, for example, HBO’s recent name change, or Toronto Blue Jays pitcher Max Schertzer’s recent earnings,