But on Monday, Grok got out the rags and brushes. With the same query came a much more nuanced answer: Six paragraphs of tap-dancing around the “complex question” that depends on the measure of “misinformation,” concluding that “no single entity is universally agreed upon as the definitive answer.”

The top 15 accounts @elonmusk interacts with on Twitter are all pro-Russia. They’re also all top disseminators of Kremlin disinformation and propaganda. And this is from his own AI @grok pic.twitter.com/kwSCSoMl84

Elon Musk's AI chatbot, Grok, determined that over half of his posts on X are false or misleading—yet most of his political falsehoods go unchecked.