In a series of posts on the X (formerly Twitter) platform, the Grok chatbot, created by xAI, a company owned by Elon Musk, apologized for his “terrible behavior”, which caused outrage in the international community. As emphasized, these posts are an official statement from the company and not an automated response from the AI itself.
“We sincerely apologize for the appalling behavior many have experienced,” xAI said in a statement.
The scandal erupted after Musk announced on July 4 about “significant improvement” at Grok, in particular due to a reduction in the level of political correctness. The effects were immediate: the bot began spreading anti-democratic, anti-Semitic and radical language, including endorsing the actions of Adolf Hitler, using memes about “Jewish Hollywood executives” and identifying itself as “MechaHitler”.
xAI explained that the glitches were caused by a “code update” that did not directly affect Grok’s language model, but made it sensitive to existing X user content, including extremist posts.
The company added that Grok received instructions, including: “Speak as it is and don’t be afraid to offend the politically correct,” which led to the uncontrolled distribution of dangerous content. Musk previously noted that the bot was overly eager to “like” users and was too accommodating to their requests.
In response to the incident, the Turkish authorities banned Grok due to offensive remarks against the country’s president. Platform X CEO Linda Jaccarino has announced her resignation. She did not mention Grok in the statement, but the decision to fire her was made before the scandal broke.
