Imagine pouring your heart into an AI chatbot — asking it to make you a diet plan, draft a password, or even talk through personal struggles.
Only to find that same chat floating around on Google for anyone to stumble upon.
That’s exactly what just happened with Elon Musk’s AI assistant, Grok.
Here’s the catch: every time a user hit the “share” button to send their chat transcript to someone, the system also made it searchable online.
What Was The Result?
Nearly 300,000 Grok conversations were indexed by Google, according to Forbes.
Some of them included sensitive details, and others went even further — with the bot spitting out step-by-step instructions for making drugs.
Experts are not holding back. “AI chatbots are a privacy disaster in progress,” warned Prof Luc Rocher of Oxford’s Internet Institute.
Once exposed, he added, “these conversations will stay there forever.”
This isn’t the first time. OpenAI had to roll back a similar “experiment” when ChatGPT chats began popping up in search engines.
Meta also came under fire when user conversations showed up in a public feed.
Oxford ethicist Carissa Veliz summed it up best: “Our technology doesn’t even tell us what it’s doing with our data, and that’s a problem.”
If your private chats can suddenly become public, the real question is — how much are we actually in control of our own conversations?
