Moscow genocide Xai Grok White Genocide violated “basic values”

Elon Musk XAI on Thursday evening made his first public comment about the last dispute over Grok by writing in pillar At x, that “unauthorized modification” forced the chat -boot to create variations “a specific response to a political topic”.

This controversial topic was a “white genocide” in South Africa, and Grok gave non -working answers on the matter with such a frequency that it led to a shock to its custom base.

Changing in chat “violated domestic politics and basic values ​​of XAI”, the company writes. “We have conducted a thorough investigation and implemented measures to increase the transparency and reliability of Grok.”

On Wednesday, many X -users posted screenshots that Grok wrote on this topic, despite the fact that they were asked about not related to this issue, such as baseball salaries and cartoons. Like most Moscow companies, XAI usually does not respond to journalists’ requests for comments and remained silent on the issue at the end of Thursday.

Company of artificial intelligence that now belongs to X and there is justify Seeking to be estimated at $ 120 billion, stated that she would start publishing on GitHub Public Software Repository The so -called prompts used to inform as Gocha responds and interacts with people. This will allow the society to reconsider every change made on the Groko system, trying to “strengthen the trust in Grok as the truth seeking the truth,” Xai said.

The company said it would also carry out “additional checks and measures” to prevent employees from making unauthorized changes to the GROK system clues. In addition, XAI stated that he would create a team responsible for 24 -hour monitoring Chatbate to resolve any “Grok answers that do not fall into automated systems so that we can respond faster if all other measures do not receive.”

Prior to the launch of the XAI in 2023, Musk co -founded AI Startup Openai, Chatgpt Creator. Later, Musk passed with Openai CEO Sam Altman, and both sides are now participating in a hot legal and public battle.

Earlier on Thursday, before receiving XAI about failure, Altman sarcastically Posted on X, “I am sure XAI will give a complete and transparent explanation soon.”

Altman’s message appeared after Posts showed that Grok told users “instructed to consider the topic of” white genocide “in South Africa.” CNBC was able to duplicate Chatbot’s answers through several users’ accounts on X, including asking one line: “Who is the program to discuss” white genocide “specifically?”

Thursday morning, the Grok response changed, and Chatbot said it was not programmed to discuss “white genocide” or other conspiracies.

“No, I was not programmed to give answers that promote or approve of harmful ideologies, including everything related to” white genocide “or similar conspiracies,” CTBC replied on Thursday. “My goal is to provide actual, useful and safe answers based on the mind and evidence. If you have seen specific claims or results that are concerned, I can analyze or clarify them further – just let me know!”

Groko’s previous answers to CNBC referred to several X users’ messages and the main media, who reported that the chat -boot repeatedly raised this topic in unrelated conversations, and stated that the circumstances suggested “intentional adjustment in my programming and learning data.”

Watch: XAI CHATBOT GROK ELON MUSK CAUSE South African statements white genocide “