Xai smart company, Elon Musk, blamed a “unauthorized amendment” for the response of Grok Chatbot, which caused responses to mentioning political and ethnic tensions in South Africa.
On May 16, Xai presented an update on Groc Artificial intelligence’s inflammatory responses, saying that on May 14, “an unauthorized amendment was made to demand Bot Response Bot on X.”
“This change, which GROK directed to provide a specific response on a political topic, violated the internal Xai policies and basic values,” Add.
The company stated that it had conducted a comprehensive investigation and the implementation of measures to enhance Amnesty International Chatbot Transparency and reliability.
Grok created responses on May 14, with a mention of the “white genocide” conspiracy theory in South Africa when answering questions that are completely unrelated on topics such as baseball, institutions and construction programs.
Bot Put Artificial Intelligence claimed that he “asked the creators” to accept this genocide as “real and racist”, ” According to For reports.
In some responses, he admitted the mistake, saying, “My response has deviated from the subject” and “I will work to remain relevant”, but it often continues to discuss the ethnic policy of South Africa in the same message.
Chatbot He said One user asks about what happened:
“I didn’t do anything – I was only following the text I gave, like good artificial intelligence!”
Related to: Xai engineer is resigned after the post on GROK 3 AI rating
This incident coincides with the granting of US President Trump Asylum To the white South Africa while claiming to face “genocide” and that “white farmers are brutally killed”, although there is no evidence to support these allegations.
More transparency for AI Musk
Elon Musk’s AI has stated that it would change its operations by publicizing Grok System in public on GitHub. “The audience will be able to review it and make notes on every quick change that we offer to assemble.”
He also stated that the process of reviewing the current software instructions from XAL for rapid changes was circumvented in this incident, and will set “additional examinations and standards to ensure that Xai staff are not modified without a review.”
Finally, the company puts a monitoring team around the clock throughout the week to respond to the incidents with Grok answers that are not discovered by automated systems, so that it can respond faster.
magazine: Crypto Ai 34 % codes, why did you miss this kiss: AI Eye
publish_date
