In a move that solidifies his war on established information platforms, Elon Musk officially announced the launch of Grokipedia , an AI-powered online encyclopedia developed by his company xAI. The platform, which quietly went live on Monday, is being positioned as a direct challenger to Wikipedia, a site Musk has frequently derided as a bastion of “left-leaning propaganda.”
However, the ambitious “truth-seeking” project immediately stumbled, facing a technical crash within an hour of launch and widespread criticism for a glaring irony: many of its articles appear to be directly adapted from Wikipedia itself.
Musk announced the launch on his social media platform, X, in his typical fashion. “Grokipedia.com version 0.1 is now live,” he posted Tuesday morning. “Version 1.0 will be 10X better, but even at 0.1 it’s better than Wikipedia imo.”
He followed up by stating the platform’s lofty ambitions: “The goal of Grok and Grokipedia.com is the truth, the whole truth and nothing but the truth. We will never be perfect, but we shall nonetheless strive towards that goal.”
The launch is the culmination of years of public complaints from Musk about the Wikimedia Foundation’s flagship project. He has repeatedly accused the volunteer-edited encyclopedia of harboring systemic ideological bias, once calling for its defunding. The project’s genesis was reportedly sparked in September by a post from David Sacks, an investor and AI adviser in the Trump administration, who called Wikipedia “hopelessly biased,” to which Musk replied, “We are building Grokipedia @xAI.”
An AI Oracle vs. Human Consensus
At the heart of Grokipedia is a fundamentally different philosophy from its rival. Wikipedia is famously built on a decentralized model of human collaboration. Millions of volunteer editors write, debate, and cite sources, with all changes and discussions logged in a transparent, public revision history.
Grokipedia is its antithesis: an automated, centralized oracle.
There is no “edit” button for users. Instead, the platform’s approximately 885,000 launch articles are generated, synthesized, and “fact-checked” by Grok, xAI’s proprietary large language model (LLM). This AI is the same one that powers the Grok chatbot on X, which notably has real-time access to data from the social media platform, allowing it to in theory update articles on breaking events faster than Wikipedia’s human-gated process.
According to Musk, users will eventually be able to “ask Grok to add/modify/delete articles,” with the AI either executing the request or explaining why it won’t. This effectively replaces Wikipedia’s communal “Talk” pages with a user-to-AI reporting system, leaving the final editorial authority in the hands of an opaque algorithm.
The ‘Ghost in the Machine’
While Musk pitched Grokipedia as a ground-up alternative, users almost immediately discovered the ghost of its predecessor in the machine.
On numerous Grokipedia entries, including articles for “Nobel Prize” and “Monday,” a small disclaimer at the bottom of the page reads: “The content is adapted from Wikipedia, licensed under Creative Commons Attribution-ShareAlike 4.0 License.”
The admission that Grokipedia is, in part, built upon the very data it claims to be replacing, became the immediate focus of online ridicule. Critics pointed out the hypocrisy of launching an “unbiased” alternative that was trained on and, in some cases, directly lifting from the source Musk decries as “propaganda.”
Musk, who has previously admitted that Grok’s training data included Wikipedia, had acknowledged the issue earlier this month in a reply on X, stating, “I know. We should have this fixed by the end of the year.”
Whose Truth? A New Bias Emerges
Despite Musk’s promise to “purge out the propaganda” before launch, critics charge that Grokipedia has not eliminated bias but simply swapped one for another. Where Wikipedia is accused of a “left-leaning” institutional bias, Grokipedia is being described as a clear reflection of Musk’s own right-leaning, culture-war-focused worldview.
The most-cited examples include:
On Gender: Grokipedia’s article begins, “Gender refers to the binary classification of humans as male or female based on biological sex.” This is a stark contrast to Wikipedia’s opening, which defines gender as “the range of social, psychological, cultural, and behavioral aspects” of being a man, woman, or other identity.
On the Jan. 6 Capitol Riot: The Grokipedia entry reportedly gives credence to false claims of “widespread claims of voting irregularities” and suggests that mainstream accounts “exaggerated” the severity of the attack.
On Elon Musk: Perhaps most telling is the platform’s entry on its own creator. Described by media outlets as “breathless” and “glowing,” the nearly 11,000-word article reads like a corporate press release. It omits significant controversies detailed on Musk’s Wikipedia page and concludes not with a neutral “Accolades” section, but with one titled “Recognition and Long-Term Vision.”
This “black box” approach, where an AI’s sourcing and reasoning are hidden, is a chief concern for experts. Without Wikipedia’s transparent edit history, it’s impossible for a user to know why the Grok AI defined a term in a specific way or chose to omit a key controversy.
Wikipedia Responds: ‘Knowledge is Human’
The Wikimedia Foundation, the non-profit that operates Wikipedia, issued a statement that subtly underscored the irony of xAI’s launch.
“Wikipedia’s knowledge is and always will be human,” the statement read. “Through open collaboration and consensus, people from all backgrounds build a neutral, living record of human understanding.”
In a direct jab at Grokipedia’s reliance on its content, the foundation added: “This human-created knowledge is what AI companies rely on to generate content; even Grokipedia needs Wikipedia to exist.”
Wikipedia co-founder Jimmy Wales, a frequent target of Musk’s criticism, has also expressed skepticism, telling The Washington Post last week that he did not have high expectations for the project, noting that AI language models “aren’t sophisticated enough” and are prone to “a lot of errors,” a phenomenon known as algorithmic hallucination.
As Grokipedia struggles to stabilize after its rocky v0.1 launch, it remains less a functional encyclopedia and more a bold, ideological statement. It has successfully created an alternative to Wikipedia, but whether it has created a less biased one is now the central, and most doubtful, question.
