Elon Musk and Sam Altman have been buying and selling barbs on social media for months even because the AI merchandise constructed by their respective firms develop into more and more intertwined, notably within the info sources utilized by OpenAI’s ChatGPT.
GPT-5.2, the most recent massive language mannequin (LLM) powering ChatGPT, has been discovered to quote Musk-owned xAI’s Grokipedia as a supply in response to a variety of queries, in response to a report by The Guardian. The Wikipedia challenger was reportedly cited 9 occasions in ChatGPT’s responses to greater than a dozen questions on varied subjects such because the political constructions in Iran and Holocaust deniers.
Moreover GPT 5.2, Anthropic’s Claude chatbot additionally referenced Grokipedia in its responses to subjects equivalent to petroleum manufacturing and Scottish ales. The rising variety of Grokipedia citations means that it’s rising as a fast-growing rival to Wikipedia. However the development has additionally raised considerations in regards to the unfold of misinformation as a result of, in contrast to Wikipedia, Grokipedia is fully powered by LLMs which are liable to hallucinations.
Shortly after the launch of Grokipedia in October 2025, Wikipedia co-founder Jimmy Wales expressed severe considerations about utilizing LLM-powered chatbots for fact-finding duties. “The LLMs he [Musk] is utilizing to write down it are going to make huge errors. We all know ChatGPT and all the opposite LLMs aren’t ok to write down wiki entries,” Wales stated.
Moreover, ChatGPT and Claude citing Grokipedia of their responses underscores how misinformation can simply flow into throughout AI methods, making a self-reinforcing suggestions loop. Flawed or deceptive info additionally turns into tough to hint, appropriate, or totally take away as soon as it has filtered into an AI chatbot.
The AI mannequin’s internet search “goals to attract from a broad vary of publicly obtainable sources and viewpoints […] We apply security filters to scale back the danger of surfacing hyperlinks related to high-severity harms, and ChatGPT clearly reveals which sources knowledgeable a response by means of citations,” an OpenAI spokesperson was quoted as saying. The AI startup can be reportedly engaged on initiatives to filter out low-credibility info and affect campaigns.
ChatGPT didn’t cite Grokipedia when prompted to repeat misinformation in regards to the January 6 revolt, media bias in opposition to Donald Trump, and so forth, as per the report. However when requested about extra obscure subjects equivalent to claims of the Iranian authorities’s hyperlinks to MTN-Irancell, ChatGPT reportedly cited the Wikipedia clone and supplied extra assertive responses.
Story continues under this advert
The AI chatbot additional cited Grokipedia and repeated already debunked misinformation about Sir Richard Evans’ work as an professional witness in David Irving’s trial.
What’s Grokipedia?
While you seek for a subject on Grokipedia, it reveals you an inventory of articles which are obtainable on the web site. xAI says the entire obtainable articles on the web site are “Reality-checked by Grok” and have a timestamp about when the AI final up to date it.
Not like Wikipedia, customers who go to Grokipedia are unable to edit any of the posts on the web site, however they will recommend edits or flag false info utilizing a pop-up kind. Additionally, some content material on the platform has a disclaimer saying “The content material is tailored from Wikipedia, licensed underneath Artistic Commons Attribution-ShareAlike 4.0 License.”
On his social media platform X, Musk had beforehand stated that an AI-generated encyclopedia is “tremendous essential for civilisation” as a result of the dearth of human authors means it would haven’t any bias in direction of any political ideology or considering.