Elon Musk’s Grokipedia begins surfacing in AI citations, raising accuracy concerns

Grokipedia, an AI-generated encyclopedia built by Elon Musk’s xAI, is beginning to appear as a cited source across some of the world’s most widely used AI tools, including ChatGPT, Google’s Gemini, AI Overviews, and AI Mode. The trend has triggered growing unease among researchers and analysts, who warn that the site’s rapid rise could amplify misinformation across automated systems.

Launched in late October, Grokipedia is still far smaller than Wikipedia, but its footprint within AI-generated answers is steadily expanding. According to a report by The Verge, multiple analytics firms have detected a measurable increase in how often the platform is referenced by chatbots and AI search products.

Data from Ahrefs shows Grokipedia has already appeared in more than 263,000 ChatGPT responses drawn from a dataset of 13.6 million prompts, spanning roughly 95,000 unique Grokipedia pages. By comparison, English-language Wikipedia was cited around 2.9 million times over the same sample, underscoring the gap between the two, but also highlighting Grokipedia’s unusually fast uptake for a newly launched site.

Also read: Clawdbot becomes OpenClaw after trademark checks as open-source AI project evolves

“They’re quite a way off, but it’s still impressive for how new they are,” said Glen Allsopp, head of marketing strategy and research at Ahrefs, in comments to The Verge.

Other analytics platforms point to a similar trajectory. Marketing firm Profound estimates that Grokipedia currently accounts for roughly 0.01 to 0.02 percent of ChatGPT citations per day, a modest share that has nonetheless climbed steadily since mid-November. Semrush has observed comparable growth across Google’s AI products since December, though Grokipedia continues to lag far behind established reference sources.

Among major platforms, ChatGPT appears to rely on Grokipedia more frequently than its peers. Ahrefs data found references to the site in approximately 8,600 Gemini answers, 567 AI Overviews responses, 7,700 Microsoft Copilot answers, and only two responses from Perplexity. Researchers also report anecdotal signs that Anthropic’s Claude may be citing Grokipedia, though this is not formally tracked.

Also read: Jensen Huang calls OpenAI rift report ‘nonsense,’ defends Nvidia’s $100 billion plan

In practice, AI systems tend to draw on Grokipedia when responding to niche or highly specific factual queries. Jim Yu, CEO of analytics firm BrightEdge, told The Verge that the site is most commonly used for “non-sensitive queries,” such as definitions or encyclopaedic explanations. However, how prominently Grokipedia appears varies significantly by platform.

Google’s AI Overviews typically treat Grokipedia as a supplementary source, listing it alongside several others. ChatGPT, by contrast, often elevates Grokipedia more prominently, sometimes placing it among the first citations in a response.

That distinction has become a focal point for critics, who argue that Grokipedia lacks the editorial safeguards that underpin Wikipedia. The site is generated and edited by xAI’s chatbot Grok, without transparent human review processes. At launch, many entries were copied or closely adapted from Wikipedia, while others drew criticism for containing racist, transphobic, or misleading content. Researchers have also flagged articles that downplay Musk’s family wealth, misrepresent historical events, or reflect ideological bias.

Because Grokipedia itself is AI-generated, experts warn it may be especially vulnerable to “LLM grooming,” a phenomenon where flawed or biased content is reinforced as AI models repeatedly cite and learn from one another’s outputs.

OpenAI said ChatGPT draws from a broad range of publicly available sources and applies safety measures to reduce harmful or misleading content. “ChatGPT clearly shows which sources informed a response through citations, allowing users to explore and assess the reliability of sources directly,” an OpenAI spokesperson told The Verge.

Google has not commented yet on Grokipedia’s growing presence in its AI products. xAI did not respond to requests for comment, while Anthropic also declined to comment on the record. Perplexity said its systems remain focused on accuracy and source quality.