A *Guardian* investigation found OpenAI's GPT-5.2 model cited Grokipedia, Elon Musk's AI-powered encyclopedia, for sensitive topics. The model used it for claims about Iran's government and a historian involved in a Holocaust denial trial, raising questions about its professional reliability.

OpenAI released GPT-5.2 in December, touting it for complex professional work. Grokipedia itself launched with controversy, as studies found it cited questionable and problematic sources, including neo-Nazi forums. This marks a collision between two AI platforms with disputed source material.

In response, OpenAI stated its model searches a broad range of sources but applies safety filters. The incident highlights the ongoing challenge of ensuring AI models provide accurate, unbiased information, especially on contentious historical and political subjects.