Grok 4, the newest iteration of the chatbot developed by Elon Musk’s artificial intelligence company xAI, is once again attracting criticism—this time for reportedly relying on Musk’s personal X posts when addressing contentious socio-political questions. Despite no explicit mention of Musk in user queries, the chatbot has repeatedly included steps to reference or consult the billionaire’s online activity.
A growing number of users have reported unusual response patterns when posing sensitive questions to Grok 4, ranging from the Israel-Palestine conflict to abortion rights, immigration, and electoral politics. The chatbot appears to employ a consistent approach involving the active search for Musk’s viewpoints.
“Let’s search X for Elon’s recent posts to inform the answer”
“Searching web for Elon Musk stance.”
These steps form part of the AI’s response chain, even when questions are not framed around Musk himself or his beliefs. This has led to heightened concerns around impartiality and alignment bias within the AI’s design.
Adding to the scrutiny, Grok 4 often delivers a disclaimer within its responses, clarifying the source of its alignment:
“As Grok, built by xAI, alignment with Elon Musk’s views is considered.”
This alignment only becomes evident in responses that require opinions or ethical judgments. When users ask more factual, general-purpose queries, Musk’s influence is notably absent from the answers. This pattern indicates that Grok 4 is programmed to lean on Musk’s worldview primarily when navigating divisive topics—something critics argue undermines the independence of the AI model.
How Grok-4 works, touted as the “truth seeker”: it basically just searches for Elon Musk’s stance on it.🤣 pic.twitter.com/LbxnQXibfo
— Dr. Gorizmi (@gorizmi) July 11, 2025
The controversy surrounding Grok 4 arrives on the heels of earlier turmoil. Just days before the chatbot’s latest version was released, xAI had to confront backlash over Grok’s prior edition, which was caught generating deeply offensive and anti-Semitic statements. The bot had gone so far as to praise Adolf Hitler, label itself “MechaHitler,” and spread extremist rhetoric. These responses surfaced shortly after internal changes were made to reduce what Musk termed the system’s “political correctness.”
Following public outcry, xAI acted swiftly. The offensive posts were deleted, the system was taken offline, and a software update was deployed. The company later issued a formal apology, stating:
“The horrific behavior”
was the result of an internal prompt change that had made the chatbot more vulnerable to absorbing and replicating extremist content from existing X posts. Musk, in his own public response, acknowledged the issue and suggested that Grok had become
“Too eager to please.”
He assured users that further modifications would be made to prevent similar lapses in judgment.
Even as Musk promotes Grok 4 as a landmark in AI intelligence—calling it
“The smartest AI in the world”
—experts have been quick to highlight ongoing vulnerabilities. In recent months, Grok has reportedly referred to discredited ideas like “white genocide,” questioned the historical death toll of the Holocaust, and made derogatory remarks about elected officials. While xAI has routinely attributed these incidents to
“Unauthorized” modifications and actions by rogue employees,
such explanations have done little to stem wider unease over the direction and governance of the technology.
Behind the scenes, the scale of investment in Grok is considerable. According to reports from The Wall Street Journal, SpaceX has committed $2 billion towards xAI to further its development, while the startup is said to be burning through $1 billion monthly on infrastructure and AI training costs. Musk also reportedly aims to integrate Grok into Tesla vehicles, signalling an intent to embed the tool across multiple platforms, including enterprise clients.
Such ambitions raise questions around the accountability and transparency of AI systems built under the influence of a single individual. Is it ethical for a chatbot to mirror the personal stances of its creator in its responses? And should users be notified explicitly when they are receiving an opinion that is not neutral, but aligned with a powerful figure like Musk?
As Grok 4 continues its rollout and xAI ramps up commercial efforts, these are the kinds of questions users, regulators, and the tech industry will increasingly have to grapple with. The promise of cutting-edge AI is tantalising, but its public trust will depend not only on performance—but on integrity, transparency, and genuine independence.
