XAI blamed an “unauthorized modification” for an error in her Grok chatbot with AI that made Grok repeatedly referred to “white genocide in South Africa” when he finds out in certain contexts in X.
On Wednesday, Grok begged responding to doses of X publications with information about white genocide in South Africa, only in response to unrelated subjects. The strange response of the votes of the X account for Grok, which responds to users with publications generated by AI when a person label “@grok”.
According to a Thursday publication of the XAI official account, a change was made on Wednesday morning to the instructions of the Grok Bot system, the high -level instructions that guide the behavior of the bot, which directed Grok to provide a “specific response” on a “political issue.” XAI says the adjustment “violated [its] Internal policies and central values ”, and that the company has carried out an exhaustive investigation.”
It is the second time that Xai recognizes and the unauthorized change in the Grok Code caused the AI to respond in a controversial way.
In February, Grok censored the little flattering mentions of Donald Trump and Elon Musk, the billionaire founder of XAI and owner of X. Igor Babuschkin, a protagonist of XAI engineering, said that Grok had been sources that strive for erroneous information, and that XAI revoked the change as soon as the users began to point it out.
Xai said Thursday that he will make several changes to prevent similar incidents from happening in the future.
As of today, XAI will publish the indications of the Grok system in Github and Changelog. The company says it
Despite Musk’s frequent warnings about the dangers of AI that were not controlled, XAI has a scarce IA security history. A recent report discovered that Grok would undress photos of women when Hen asked. Chatbot can also be considerably more rude than AI like Gemini and Google Chatgpt, cursing without much reverse to speak.
A study by Safeai, a non -profit organization with the aim of improving the responsibility of AI Labs, found that Xai occupies a bad place in safety among its peers, due to its “very weak” risk management practices. Earlier this month, XAI lost a self -imposed deadline to publish a finished security framework.
