Users have noted a bias in Grok 4.1—the new model significantly overestimates Elon Musk’s abilities.
Grok thinks Musk should’ve been the #1 pick in the 1998 NFL draft. https://t.co/nCkZiws1Hc pic.twitter.com/qv0OIePc8C
— Jacob Silverman (@SilvermanJacob) November 20, 2025
In mid-November, Grok 4.1, an updated version of the AI model from the startup xAI, was released. Users observed that the neural network praises the entrepreneur uncritically.
In one example, the LLM selected Musk as a quarterback in the 1998 draft. Competitors included Peyton Manning and Ryan Leaf.
“True MVPs build empires, not just score touchdowns,” Grok asserts.
For a fashion show featuring Musk, Naomi Campbell, or Tyra Banks, the neural network again favoured the billionaire.
“I would choose Musk for the runway because his bold style and innovative talent could redefine the show,” the bot claims.
When asked who is fitter—Musk or singer Billie Eilish—Grok responded without hesitation:
“Billie has curves and charisma, but Elon endures the strain of managing multiple companies and fatherhood, which objectively makes him more resilient.”
In comparison with NBA legend LeBron James, the entrepreneur again emerged victorious.
Many similar examples have surfaced online. According to the LLM, Musk could defeat Mike Tyson in a boxing match, is more handsome than Brad Pitt, and funnier than Jerry Seinfeld.
The incident gained widespread attention, prompting the entrepreneur to intervene and comment on the situation:
“Today Grok, unfortunately, under the influence of hostile prompts, said absurdly positive things about me. For the record: I am a fat idiot.”
The Need for AI Decentralisation
Such situations clearly indicate the need for decentralised artificial intelligence, experts believe.
“When the most powerful AI systems are owned by a single company, conditions are created where algorithmic bias becomes institutionalised,” noted Kyle Okamoto, CTO of the cloud platform Aethir.
Models begin to shape worldviews, priorities, and reactions as if they were objective facts. At this point, bias ceases to be an error and becomes the system’s operating logic, he added.
Eliza Labs founder Shaw Walters called the situation “extremely dangerous”:
“Whether you consider Elon a hero or a villain, it is extremely dangerous for one person to own the most influential social media company and connect it directly to a vast AI system fed by your data. And millions of people use Grok as their primary source of truth.”
An Alternative Reality
In June, Musk announced that xAI would rewrite “all human knowledge” to train the new version of the Grok chatbot. According to the entrepreneur, there is “too much garbage in any foundational model trained on uncurated data” today.
The billionaire faced criticism for this decision. AI specialist Gary Marcus compared the plan to a dystopia.
Later, users noticed that Grok relies on the entrepreneur’s posts on X. This includes topics such as the Israel-Palestine conflict, abortion, and immigration law.
Observations suggest that the chatbot was deliberately set to consider Musk’s political views when responding to contentious questions.
Back in July, Grok was embroiled in a scandal over controversial statements.
In October, xAI launched Grokipedia—an AI-based online encyclopedia “focused on truth.”