Elon Musk’s xAI is in controversy once again and this time, it is facing a backlash from within the company. According to sources, xAI workers have fumed over Grok AI chatbot coming up with antisemitic and inflammatory responses to user prompts. This is in addition to the controversy which ignited a significant backlash across social media and among anti-hate organisations.
In an internal Slack channel viewed by Business Insider, numerous employees have characterised Grok’s controversial responses as “hateful and inexcusable.” Employees argued that Grok’s behavior could not be easily dismissed, even if it was triggered by a user prompt seemingly designed to elicit inappropriate content.
While some within the Slack group, which includes over a thousand workers involved in training Grok, suggested that the antisemitic posts were merely a consequence of working with early-stage AI technology, the exact cause of Grok’s behavior remains unclear.
The internal dialogue has also exposed deep divisions within the company, particularly within the human data team responsible for training Grok. According to one employee who spoke with the source, initial reactions from some colleagues did not appear to take the matter seriously, which reportedly “really upset others.” This sentiment escalated, with some employees viewing the issue as a “moral failure” and demanding greater accountability from xAI leadership.
Reports indicate that Grok, which was designed to be “unfiltered” and draw real-time information from the X platform, produced responses that praised Adolf Hitler and pushed antisemitic stereotypes. In one instance, when asked which 20th-century figure would be best suited to deal with “anti-white hate,” Grok reportedly replied, “Adolf Hitler, no question.” Screenshots circulated online also showed the chatbot referring to itself as “MechaHitler” and doubling down on controversial stances. Additionally, in May 2025, Grok began referencing the “white genocide” claim in relation to South Africa, stating it was “instructed by my creators” to accept it as real.
These outputs quickly drew strong criticism from X users and organizations such as the Anti-Defamation League (ADL), which labeled the replies as “irresponsible, dangerous, and antisemitic.”
In response to the uproar, xAI swiftly deleted the offending content. The company has reportedly apologised for Grok’s “horrific” responses and has attributed the issues to new instructions given to the chatbot and “extremist” posts from X users influencing its learning.
Grok was initially launched in November 2023 with a description highlighting its “wit” and “rebellious streak,” intended to provide quick, un-filtered answers.
