ChatGPT and Google's AI tools increasingly cite Elon Musk's Grokipedia, according to new data analysis. The AI-generated encyclopedia appears in hundreds of thousands of chatbot responses despite accuracy concerns.
Grokipedia citations appeared in more than 263,000 ChatGPT responses from 13.6 million prompts tested, according to Ahrefs research. The platform launched in late October as part of Musk's xAI ecosystem and now accounts for 0.01 to 0.02 percent of daily ChatGPT citations.
Google's AI Overviews, AI Mode, and Gemini also reference Grokipedia, with similar increases detected since December. Microsoft Copilot shows around 7,700 citations and Perplexity shows around 2 citations, according to Ahrefs data, while anecdotal reports suggest Anthropic's Claude may also use the source.
Analysts found Grokipedia appears most frequently for niche, obscure, or highly specific factual questions. BrightEdge CEO Jim Yu told The Verge that ChatGPT and AI Overviews use it primarily for "non-sensitive queries" like encyclopedic lookups and definitions.
The platform differs fundamentally from Wikipedia in its lack of human oversight. Grokipedia is produced and edited by xAI's chatbot Grok, making it susceptible to what researchers call "LLM grooming" or data poisoning.
"Grokipedia feels like a cosplay of credibility," said Leigh McKenzie, director of online visibility at Semrush. "It might work inside its own bubble, but the idea that Google or OpenAI would treat something like Grokipedia as a serious, default reference layer at scale is bleak."
Trinity College Dublin's Taha Yasseri warned that "fluency can easily be mistaken for reliability" when AI systems cite AI-generated content. Unlike Wikipedia's transparent human editing process, Grokipedia relies on opaque sourcing that includes personal websites and blog posts.
OpenAI spokesperson Shaokyi Amdo said that ChatGPT utilises a variety of sources that are all accessible to the public and includes citations to make users aware of their reliability. Google declined to comment on its AI tools' use of Grokipedia.
Despite the increasing citations, Grokipedia remains a minor source compared to established references. English-language Wikipedia appeared in 2.9 million ChatGPT responses during the same testing period, nearly 11 times more than Grokipedia.
The trend raises broader questions about AI systems reinforcing biases and errors through circular sourcing. As AI-generated content proliferates, the risk of misinformation spreading through automated citation chains grows proportionally, highlighting ongoing tensions in the AI industry's relationship with Musk's ventures.
