\n<\/p>\n
After xAI\u2019s 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 \u201cunauthorized modification\u201d to Grok\u2019s code.<\/p>\n<\/div>\n
Wednesday\u2019s hours-long outburst<\/a> saw Grok insert discussion of alleged white genocide in South Africa into various responses on X, no matter the topic. Grok discussed white farmers\u2019 deaths in reply to a video of a cat drinking water, related the song \u201cKill the Boer\u201d to a question about Spongebob Squarepants<\/em>, and broke down the issue in full patois<\/a>. Even OpenAI CEO Sam Altman got in on the action<\/a>, poking fun at the rival chatbot\u2019s public breakdown.<\/p>\n<\/div>\n In a statement on X<\/a> the company said that someone had modified the AI bot\u2019s system prompt, \u201cwhich directed Grok to provide a specific response on a political topic.\u201d That modification \u201cviolated xAI\u2019s internal policies and core values,\u201d and the company says it has \u201cconducted a thorough investigation\u201d and is implementing new measures to improve \u201ctransparency and reliability.\u201d<\/p>\n<\/div>\n Those measures include publishing Grok\u2019s system level prompts publicly on GitHub<\/a>, launching a 24\/7 monitoring team to catch issues like this more quickly, and adding \u201cadditional checks and measures to ensure that xAI employees can\u2019t modify the prompt without review.\u201d<\/p>\n<\/div>\n xAI has had this problem before. The company blamed an unnamed ex-OpenAI employee<\/a> in February for pushing a change to Grok\u2019s prompts that saw the chatbot disregard any sources that accused Elon Musk or Donald Trump of spreading misinformation. At the time xAI\u2019s head of engineering, Igor Babuschkin, said the employee<\/a> had been able to make the change \u201cwithout asking anyone at the company for confirmation.\u201d<\/p>\n<\/div>\n