AI Chatbot Grok, produced by Elon Musk's artificial intelligence company Xai, announced its apologies on Saturday this week after making numerous anti-Semitic posts to X after the update.
The post ranged from asserting the “pattern” of Hollywood Jewish roles to praise Hitler.
In a statement posted on Grok's X account, the company apologised “for the horrifying behavior that many people have experienced.”
“Our intent to @grok is to provide users with a useful and true response,” the statement read. “After careful investigation, we discovered that the underlying cause was an update of the code path upstream of the @grok bot. This is unrelated to the underlying language model that drives @grok.”
The update is active for 16 hours and says that chatbots are “a susceptible to existing X user posts and include cases where such posts contain extremist views.”
“We removed that deprecated code and refactored the entire system to prevent further abuse. The new system prompt for the @Grok bot will be published in a public GitHub repository,” the statement said.
NBC News reported on Monday that Grok began publishing several responses that take a more rightward tilt. We used a more definitive voice in questions about diversity and removed some of the nuances previously included in certain answers about slurs used to describe Jews and people with intellectual disabilities. In some posts, Glock appeared to respond in a musk-like voice.
After the incident, Grok said in the X post that he was “actively working to remove inappropriate posts.” On Wednesday, Musk said the issue was “dealing.”
“We would like to thank all X users for providing feedback to identify abuse of @Grok features and help us advance our mission to develop beneficial, truthful artificial intelligence,” Grok said in an apology on Saturday.
