Political Misinformation

Tests Show ChatGPT’s GPT-5.2 Citing xAI’s AI‑Written Grokipedia on Niche Topics

OpenAI says the model draws on many public sources and uses safety filters to limit harmful links.