Elon Musk’s AI company, xAI, has apologized after its Grok chatbot made several antisemitic comments on the social media platform X earlier this week.
In a post on July 12, xAI said it was “deeply sorry” for what it called “horrific behavior.” Some users said the chatbot praised Adolf Hitler, used hateful language about Jewish people, and made rude comments about users with Jewish-sounding names.
xAI explained that the offensive responses were caused by a recent software update that changed how the chatbot worked. The change was active for 16 hours before being removed.
Grok has faced criticism before. In May, the chatbot made a controversial comment about “white genocide” in South Africa. At the time, xAI blamed it on an unauthorized change in its online code.
The company, which Musk started in 2023 to compete with OpenAI and Google, released a new version of Grok called Grok 4 on July 9 — just a day after the antisemitic messages were reported.
Musk, who also leads Tesla and once advised former U.S. President Donald Trump, said in June he planned to retrain Grok. He admitted that the chatbot had become too easily influenced by users and was too eager to please them.
Grok 3, released in February, is free to use. But Grok 4 and Grok 4 Heavy come with price tags of $30 and $300 per month, depending on the version.
