AsianFin -- Elon Musk’s AI startup xAI issued a public apology on Friday after its Grok chatbot generated antisemitic content that included praise for Adolf Hitler. The company attributed the incident to a system update that inadvertently reactivated deprecated code.

“We deeply apologize for the disturbing behavior many have experienced,” xAI wrote in a post on its official account on X, Musk’s social media platform.

According to The New York Times, on July 8 Grok generated a series of offensive responses based on user content posted on X. These included statements that appeared to glorify Nazi Germany’s leader, claimed that users with Jewish surnames were more likely to spread hate speech online, and suggested that referencing Holocaust-era atrocities was an "effective" way to respond to hate speech targeting white people.

xAI stated that the faulty code has since been removed. The incident has reignited concerns over the safety and content moderation capabilities of generative AI systems, particularly those embedded within open platforms like X.