OpenAI’s latest language model, GPT‑5.2, has started using content from Grokipedia, an AI‑written encyclopedia created by Elon Musk’s company xAI. Unlike Wikipedia, which relies on human editors, Grokipedia generates and updates its articles entirely with AI. While it was designed to offer a fresh perspective, some experts worry it may spread inaccurate or biased information.
Recent tests revealed that ChatGPT often referenced Grokipedia when answering questions about less-known topics. For instance, queries on certain historical figures or obscure international events sometimes drew responses directly from Grokipedia entries. Experts say this raises questions about reliability, because unlike Wikipedia, Grokipedia does not have human fact-checkers to verify its content.
Interestingly, GPT‑5.2 seems to avoid using Grokipedia for widely publicised or controversial topics, such as major political events or widely debunked health claims. This selective referencing suggests the AI chooses sources based on context—but it also means obscure topics could carry unchecked errors.
Researchers warn that AI models relying on algorithmically generated sources may inadvertently spread misinformation, especially if users assume ChatGPT answers are always neutral and verified. While OpenAI emphasizes that its models draw from a broad range of publicly available sources and include safety filters to reduce harmful content, the Grokipedia connection highlights the ongoing challenge of balancing knowledge breadth with credibility.
For readers, this is a reminder to approach AI answers critically. Even the most advanced AI tools can reflect the quirks, biases, or errors present in their source material. As AI becomes a more common companion for information and learning, understanding where the information comes from is more important than ever.
Also Read: Canadian PM rejects China deal after Trump’s 100% tariff alert
