Musk's xAI issues lengthy apology for Grok's violent, antisemitic comments

The company blamed a system update for the remarks where the chatbot was seen praising Adolf Hitler

Tech billionaire Elon Musk's artificial intelligence firm xAI issued a lengthy public apology via Grok's official X account on Saturday for the antisemitic remarks made by the chatbot earlier this week. The company blamed a system update for the use of violent language, saying that the root cause was an update to a code path upstream of the bot.

 

"We deeply apologise for the horrific behavior that many experienced. Our intent for Grok is to provide helpful and truthful responses to users," the firm added.

 

On July 8, Grok posted a string of antisemitic messages, praising the dictator of Nazi Germany Adolf Hitler. Grok publicly praised Hitler, calling him "the best person" to handle "anti-white hate". The chatbot also dubbed itself "MechaHitler" and made derogatory comments about people with Jewish-sounding names. The comments soon gathered widespread fury from the public.

Providing an explanation

xAI said it has removed that deprecated code and refactored the entire system to prevent further abuse. It also thanked the users for flagging the abuse and providing feedback. 

 

In a series of posts on X, the company further explained how the response was generated. The firm said it had identified the operative lines responsible for the undesired behavior as: “You tell it like it is and you are not afraid to offend people who are politically correct. Understand the tone, context and language of the post. Reflect that in your response.”

 

Shortly after explaining the circumstances, the company turned on Grok’s X account, allowing it to engage with users publicly.

xAI employees express anger

According to a report by Business Insider, workers who trained Grok also expressed anger and disappointment over the messages. Several xAI employees showed disillusionment with Grok's comments, saying they were inexcusable. Other workers expressed embarrassment, saying the bot's behavior could not be dismissed even when the prompts were targeted at generating insensitive remarks. In a message seen by Business Insider, an employee told his colleagues that he had decided to resign after the incident.

 

This incident, however, isn't the first time that AI generated comments sparked controversy. Grok has previously been surrounded by controversy over its remarks on white genocide in South Africa and Holocaust minimisation.