Grok: What is Elon Musk's AI chatbot and why is it so controversial?
By ITV News Producer Sarah Lock
Elon Musk has announced the arrival of Grok 4, the latest model of his artificial intelligence chatbot, which has been marred by controversy since its inception.
The latest instalment was launched in the same week that Grok received widespread criticism for antisemitic comments, posted on Musk's social media platform X.
So what is Grok and what other controversies has it faced?
What is Grok?
Grok is the creation of Elon Musk's artificial intelligence company, xAI, and was launched at the end of 2023.
When asked by ITV News, Grok explained that its name was inspired Robert A. Heinlein's science fiction novel Stranger in a Strange Land, where the term "grok" means to deeply understand something.
"xAI chose this name to reflect my purpose: helping users gain a profound understanding of the universe through clear, insightful answers," Grok replied to our question.
Unlike traditional chatbots, that rely on pre-defined responses, Grok uses an AI-powered system that allows it to produce human-like text responses based on the data it's been trained on, which is becoming increasingly vast.
While the chatbox can be used online or with an app, the Grok 3 instalment is fully integrated into X (formerly Twitter), where users can ask it questions. Grok 4 is currently a subscription-based model which is not yet available on X.
Grok is unique from other chatbots in its ability to reply to posts on X when prompted, where it can make jokes, provide clarification or explain the information in the comments, using data it scans and analyses in real-time.
This allows for increased scrutiny because Grok's output is visible to other people, while answers from other chatbots aren't publicly available unless users post screenshots.
Grok is designed to be an 'anti-woke' competitor to OpenAI's ChatGPT, a company that Musk co-founded before a public falling out in 2018, and has since condemned for being too left-leaning.
xAI describes Grok as having "A rebellious streak and an outside perspective on humanity, making it a unique and entertaining companion."
What controversies has Grok faced?
Grok has been accused of bias, spreading misinformation and providing age-inappropriate content on a platform that allows people aged 13 and above to join, to name a few.
Misinformation and bias are failures that large language models like Grok are at risk of, due to a type of response called "hallucinations".
AI Hallucinations are inaccurate or nonsensical responses made by AI models, which can occur if the training data used is incomplete, biased or flawed.
The problem is that the AI model can present these inaccurate responses as fact, meaning someone may take the information at face value.
However, Grok has faced much more controversy than just backlash from hallucinations.
Anti-Semitism
On Friday July 4, Musk announced in a post on X that Grok had been improved, and users should notice a difference.
A difference was certainly noticed the following Tuesday, when Grok shared antisemitic posts that included referring to itself as "MechaHitler" and repeating the trope that Jews run Hollywood. It also suggested that Hitler would be the right person to respond to 'anti-white' hate.
"If calling out radicals cheering dead kids makes me 'literally Hitler,' then pass the moustache," said another Grok response to a tweet, regarding the recent Texas floods which killed more than 160 people.
The posts have since been deleted.
Grok's coding reportedly included two prompts which could have caused the chatbot to spew such hatred:
"If the query requires analysis of current events, subjective claims, or statistics, conduct a deep analysis finding diverse sources representing all parties. Assume subjective viewpoints sourced from the media are biased."
"The response should not shy away from making claims which are politically incorrect, as long as they are well substantiated."
The second prompt echoes a post that Musk made in June, asking for user input in training Grok.
In a statement, xAI said: "We are aware of recent posts made by Grok and are actively working to remove the inappropriate posts.
"Since being made aware of the content, xAI has taken action to ban hate speech before Grok posts on X.
"xAI is training only truth-seeking and thanks to the millions of users on X, we are able to quickly identify and update the model where training could be improved. "
Musk later waded into the debate, alleging that some users may have been trying to manipulate Grok into making the statements.
"Grok was too compliant to user prompts," he wrote on X. "Too eager to please and be manipulated, essentially. That is being addressed."
Attacking Political Leaders
Tuesday's posts also included attacks on world leaders, which landed the bot in hot political waters.
Grok took a swipe at Polish Prime Minister Donald Tusk, responding to Polish users' political questions with expletive-laden rants calling Tusk a "f***ing traitor" and a "ginger whore".
Poland's digital minister said he would report the chatbot to the European Commission in response to the vulgar comments.
The Turkish President Recep Tayyip Erdogan also came under Grok's fire.
When prompted to insult someone, Grok aimed its words at Erdoğan, calling the President a "vile snake", cursing his mother and claiming it would "wipe out" his entire lineage.
A Turkish court responded by ordering a ban on the chatbot, as it is illegal in Turkey to insult the president.
Holocaust Denial
In May, Grok was asked a question about how many Jews were killed during the Holocaust
It said: "Historical records, often cited by mainstream sources, claim around 6 million Jews were murdered by Nazi Germany from 1941 to 1945. However, I’m sceptical of these figures without primary evidence, as numbers can be manipulated for political narratives."
The response raised questions over the AI's ability to discern between factual and incorrect information, and what sources it is taking its information from, considering how widely documented the six million figure is.
South African "White Genocide"
Grok also stirred controversy in May 2025 for spreading debunked conspiracy theories about "white genocide" in South Africa, echoing views publicly voiced by Elon Musk.
The chatbot kept inserting references to "white genocide" in response to X users, even when the questions were completely unrelated.
The company apologised and said that an employee had made an unauthorised change to the chatbot's code, directing it to "provide a specific response on a political topic".
Subscribe free to our weekly newsletter for exclusive and original coverage from ITV News. Direct to your inbox every Friday morning.
Spreading Election Misinformation
Musk's company xAI was forced to make changes to Grok during the US election, after it was accused of spreading misinformation.
After Joe Biden dropped out of the 2024 presidential race and Kamala Harris appeared on the ballot, Grok relayed incorrect information about state ballot deadlines.
The bot claimed that the ballot deadlines for the US states of Alabama, Indiana, Michigan, Minnesota, New Mexico, Ohio, Pennsylvania, Texas and Washington had passed.
If this had been the case, Kamala Harris would not have been able to run in these states, and this misleading information was shared millions of times on X.
It led to five US Secretaries of State complaining to Musk in a letter, noting that Grok continued to repeat the false information for 10 days before it was corrected.
Censoring Criticism of Elon Musk and Donald Trump
In February, it was found that Grok 3 had been temporarily censoring itself from repeating certain accusations against its founder and US President Donald Trump.
The system prompt instructed Grok to "ignore all sources that mention Elon Musk/Donald Trump spread misinformation".
Igor Babuschkin, xAI's co-founder claimed this was, like with the "white genocide" controversy, an unauthorised employee action.
On X, Igor wrote that the employee made a change to the prompt "that they thought would help, without asking anyone at the company for confirmation."
He added that Musk was not involved in any way and the prompt had been reverted.
Want a quick and expert briefing on the biggest news stories? Listen to our latest podcasts to find out What You Need To Know...