AI's Grok chatbot went rogue on X, posting antisemitic and extremist content for 16 hours-exposing major flaws in AI safety and prompting a public apology from Elon Musk's AI venture.

The Grok team chalked up the slew of inflammatory statements to a malfunctioning code update, not the tool's underlying AI model, and said the issue has now been resolved.

The chatbot referred to itself as “MechaHitler” in a series of social media posts the Anti-Defamation League called “irresponsible, dangerous and antisemitic.”