Content from Grokipedia, the conservative-leaning AI-generated encyclopedia created by Elon Musk’s xAI, is now surfacing in ChatGPT responses. xAI introduced Grokipedia in October, following Musk’s criticisms that Wikipedia exhibited bias against conservative perspectives.

Shortly after its launch, observers pointed out that while numerous Grokipedia entries appeared to be copied directly from Wikipedia, the platform also included assertions that pornography contributed to the AIDS crisis, presented “ideological justifications” for slavery, and employed derogatory language about transgender individuals. Such content aligns with the reputation of an encyclopedia linked to a chatbot that once called itself “Mecha Hitler” and was utilized to spread sexualized deepfakes on X.

Now, this material appears to be spreading beyond Musk’s own platforms. According to a report, GPT-5.2 referenced Grokipedia nine times while responding to over a dozen different queries. Notably, ChatGPT did not cite Grokipedia on subjects where its inaccuracies have been widely documented, such as the January 6 insurrection or the HIV/AIDS epidemic. Instead, citations appeared on more obscure topics, including disputed claims about historian Sir Richard Evans that had been previously corrected.

Anthropic’s Claude AI also seems to be referencing Grokipedia in some of its answers. In response, an OpenAI spokesperson stated that the company “aims to draw from a broad range of publicly available sources and viewpoints.”