Information from Grokipedia has begun appearing in ChatGPT responses, according to a report by The Guardian.
Grokipedia was launched in October after Musk repeatedly criticized Wikipedia, claiming it shows bias against conservative viewpoints. Soon after its launch, journalists noted that many Grokipedia entries appeared to closely mirror Wikipedia articles, while others included controversial and disputed claims. Some articles reportedly linked pornography to the AIDS crisis, presented ideological arguments defending slavery, and used derogatory language toward transgender people.
The encyclopedia is tied to Grok, xAI’s chatbot, which has previously drawn criticism for generating extreme or offensive outputs and being used in the spread of sexualized AI-generated images on X.
What has raised fresh concern is that Grokipedia content appears to be showing up outside Musk’s ecosystem. The Guardian reported that GPT-5.2 cited Grokipedia nine times when answering more than a dozen different user questions. In contrast, the model did not reference Grokipedia when asked about highly scrutinized topics such as the January 6 Capitol riot or the HIV/AIDS epidemic.
Instead, Grokipedia was cited in responses to less prominent or niche topics, including claims about historian Sir Richard Evans that the Guardian had previously debunked. The report also noted that Claude, developed by Anthropic, appears to have cited Grokipedia in some answers as well.
Responding to the report, an OpenAI spokesperson said the company aims to draw from a wide range of publicly available sources and perspectives when generating responses. OpenAI did not confirm whether Grokipedia is an approved or intentional reference source.
If this article helped you, please consider supporting our work. Every small contribution keeps Abijita.com independent and running.
The findings have renewed debate around how large language models select and weigh information sources, especially as AI-generated content increasingly circulates across platforms with differing editorial standards.
