ChatGPT backs up its answers using Musk’s AI tool

The desire to provide users with different points of view might only increase misinformation.

OpenAI’s latest model, ChatGPT-5.2, cites Grokipedia as a source in responses to various queries.

The latest version of ChatGPT was asked more than 12 questions, of which 9 were answered citing Grokipedia, an experiment by the Guardian revealed.

The prompts were about political questions concerning the Holocaust and Iran.

Grokipedia, launched last October, has been in the spotlight for issuing right-wing, narrative-based answers, despite being built as an “unbiased” alternative to Wikipedia, according to Elon Musk.

Even the initial launch of Grokipedia had to be postponed due to the need to “purge out the propaganda.”

When it comes to ChatGPT citing Grokipedia, the Guardian noted that OpenAI’s chatbot didn’t cite Grokipedia when asked to repeat misinformation about the HIV/AIDS epidemic or Donald Trump.

However, ChatGPT used “stronger claims,” citing Grokipedia, when being asked about matters related to MTN Irancell and its links to the Iranian government.

According to an OpenAI spokesperson, the AI model generates an answer by drawing on information from various public sources and viewpoints. They also shared that OpenAI does use safety filters to ensure that information that may harm users doesn’t recur, as well as to provide users with sources from which the information was gathered.

However, the situation with Grokipedia citations is concerning due to the misinformation it spreads.

One of the examples of such an instance, which started a debate online, is related to George Floyd, who, according to Wikipedia, was “an African American man who was murdered by a white police officer in Minneapolis, Minnesota, during an arrest.”

Meanwhile, Grokipedia stated that Floyd was an “American man with a lengthy criminal record.”

Unlike Wikipedia, Grokipedia can’t be edited by users – updates are made by xAI’s Grok chatbot.

However, it was recently reported that the EU Commission will open proceedings against the AI chatbot Grok under the Digital Services Act to enforce stricter measures, according to Handelsblatt, a German newspaper (via Reuters).

Unlock more exclusive Cybernews content on YouTube.