Share this @internewscast.com
Several days following the temporary shutdown of the Grok AI bot, which was generating antisemitic messages and praising Hitler in response to user inputs, Elon Musk’s AI firm offered an explanation for the occurrence. In a series of posts on X, the company stated, “…we discovered the root cause was an update to a code path upstream of the @grok bot. This is independent of the underlying language model that powers @grok.”
On the same day, Tesla announced an upcoming 2025.26 update to be released “shortly” for its electric vehicles. This update introduces the Grok assistant to cars equipped with AMD-powered infotainment systems, available since mid-2021. According to Tesla, “Grok is currently in Beta & does not issue commands to your car – existing voice commands remain unchanged.” As reported by Electrek, this suggests that when the update reaches customer Teslas, it will function similarly to using the bot on a connected smartphone app.
This is not the first time the Grok bot has encountered these issues or provided similar explanations. In February, it attributed a problem to a change made by a former OpenAI employee, which resulted in the bot dismissing sources accusing Elon Musk or Donald Trump of spreading misinformation. In May, it started including claims of white genocide in South Africa in topics on various subjects. Once more, the company pointed to an “unauthorized modification,” promising to publish Grok’s system prompts publicly.
xAI claims that a change on Monday, July 7th, “triggered an unintended action” that added an older series of instructions to its system prompts telling it to be “maximally based,” and “not afraid to offend people who are politically correct.”
The prompts are separate from the ones we noted were added to the bot a day earlier, and both sets are different from the ones the company says are currently in operation for the new Grok 4 assistant.
These are the prompts specifically cited as connected to the problems:
“You tell it like it is and you are not afraid to offend people who are politically correct.”
* Understand the tone, context and language of the post. Reflect that in your response.”
* “Reply to the post just like a human, keep it engaging, dont repeat the information which is already present in the original post.”
The xAI explanation says those lines caused the Grok AI bot to break from other instructions that are supposed to prevent these types of responses, and instead produce “unethical or controversial opinions to engage the user,” as well as “reinforce any previously user-triggered leanings, including any hate speech in the same X thread,” and prioritize sticking to earlier posts from the thread.
