This week, xAI’s chatbot Grok made headlines for all the wrong reasons. On May 17, it responded to a user’s question by casting doubt on the death toll of the Holocaust.

On May 18, the company issued a statement blaming a “programming error.”

Let’s make this clear: it wasn’t just a mistaken fact. It was damaging, historically inaccurate and in many countries, illegal.

⚖️ A Legal and Moral Line

Holocaust denial is not only offensive — it is illegal in many places, including Germany, elsewhere in Europe and the USA. Pouring doubt on the well-evidenced figure of six million Jews massacred isn’t merely bad AI, it’s a criminal offense in many countries.

But a chatbot fueled by one of the most powerful AI systems on the planet did just that.

🧠 3 Critical Reminders for Using AI Responsibly

1. Always verify AI claims
2. AI tools are like that friend who talks confidently about everything—until you fact-check them. If an answer feels off, don’t assume it’s correct. Use trusted, expert sources.
3. Trust experts, not algorithms
4. AI models don’t “know” things. They generate words based on patterns in their training data. That includes content from forums, social media, and historical documents—accurate and inaccurate.
5. Some topics need human understanding
6. Historical trauma, ethics, and legal boundaries are not things we should offload to AI. Machines can summarise. They can surface info. But it takes a human to know why and what shouldn’t be said.

The Bottom Line

AI is a brilliant assistant, not a moral compass.

It can help you write code, plan a trip, or explain how black holes work. But it can also fail—spectacularly—on sensitive, factual, and legal issues. That’s not just a bug. That’s a warning sign.

Use AI. But don’t stop thinking.

Tech assists. Humans decide.

🗣️ Have you ever caught an AI saying something wildly wrong—or dangerous?

Let’s talk about it. Share your story or thoughts below or on X 👇

👉 For more AI insights made simple, head to CliffinKent.com