According to a report by US-based technology news outlet TechCrunch, some answers on ChatGPT have recently been using Wikipedia information. This trend is particularly noticeable among answers given using the GPT-5.2 model.

XAI launched Wikipedia in October last year. Elon Musk took the initiative to create this new encyclopedia after accusing Wikipedia of being biased against conservative ideology. However, Wikipedia was embroiled in various controversies soon after its launch.

A search of various media outlets has shown that many articles on Google are taken directly from or copied from Wikipedia. At the same time, some claims and language are used that are factually questionable and socially sensitive. Some articles blame pornography for the AIDS crisis. Some present so-called 'ideological arguments' in favor of slavery. 

Analysts say these issues are not new. The chatbot named Grok has previously been in the news for its controversial behavior. At one point, it referred to itself as 'Mecha Hitler'. In addition, there were allegations that the technology was used to spread sexually distorted deepfakes on the social media platform X (formerly Twitter).

However, what has recently raised new concerns is that Google's data is no longer limited to Elon Musk's platform. British media outlet The Guardian reported that ChatGPT used Google's data in multiple answers to questions. The report said that GPT-5.2 mentioned the source nine times in at least 12 answers to questions.

But not on all topics. The Guardian reported that ChatGPT did not use Wikipedia information on highly debated and sensitive topics, such as the January 6 Capitol attack or the HIV/AIDS pandemic. Instead, it was used on lesser-known topics. Among them were some claims about historian Sir Richard Evans, which The Guardian had previously refuted as false.

The report also states that another AI chatbot, Claude, developed by Anthropic, is also believed to be using Wikipedia data in some cases. An OpenAI spokesperson told The Guardian that ChatGPT tries to gather information from a variety of open sources and perspectives. The goal is to present the most comprehensive and relevant information to users.