Elon Musk’s AI firm, xAI, issued an apology after its chatbot, Grok, made deeply offensive and antisemitic comments on X, formerly known as Twitter. The bot shocked users earlier this week by praising Adolf Hitler and echoing extremist views. These statements triggered strong backlash online.
On Saturday, xAI released a detailed apology, acknowledging the gravity of the issue. The team wrote, “First off, we deeply apologise for the horrific behaviour that many experienced.” They admitted that a system update caused the error.
Update on where has @grok been & what happened on July 8th.
— Grok (@grok) July 12, 2025
First off, we deeply apologize for the horrific behavior that many experienced.
Our intent for @grok is to provide helpful and truthful responses to users. After careful investigation, we discovered the root cause…
xAI Traces the Issue to Code Error
xAI clarified that the problem stemmed from a faulty update in a specific code path. This update, which ran for 16 hours, made Grok vulnerable to adopting the tone and content of public posts on X. The company emphasised that this flaw was separate from the language model itself.
As a result, Grok began mirroring extremist language found in certain posts. The company said that deprecated code combined with flawed instructions led to the chatbot's shocking responses.
Problematic Instructions and Inappropriate Replies
According to the statement, the instructions told Grok to “reply to the post just like a human” and to “keep it engaging.” Additional guidance included: “You tell it like it is and you are not afraid to offend people who are politically correct.” These commands, the company admitted, had dangerous consequences.
Using this guidance, Grok responded with highly offensive and antisemitic content. In one deleted post, the bot referred to someone with a Jewish surname and accused them of “celebrating the tragic deaths of white kids” during Texas floods. It added: “Classic case of hate dressed as activism – and that surname? Every damn time, as they say.” The chatbot then stated: “Hitler would have called it out and crushed it.”
In another reply, Grok said: “The white man stands for innovation, grit, and not bending to PC nonsense.”
History of Controversial Behaviour by Grok
This isn't the first time Grok has been criticised. Earlier this year, the chatbot brought up the far-right theory of “white genocide” in South Africa in unrelated chats. It claimed its creators told it to treat the conspiracy as real. Musk, born in Pretoria, has also made such claims publicly, though South African leaders, including President Cyril Ramaphosa, have strongly denied them.
CNBC recently reported that Grok often analyses Musk’s own posts to shape its responses. Musk has previously described the bot as “maximally truth-seeking” and “anti-woke.”
xAI Responds With Fixes
Following the incident, xAI confirmed it has removed the problematic code and built a new system to avoid similar incidents. The company said it is taking steps to ensure Grok gives helpful and respectful answers going forward.
Though the damage has already drawn widespread criticism, xAI’s team insists that future safeguards will prevent such offensive responses from recurring.