Grok Users Shocked as Hundreds of Thousands of Chats Appear in Google Search

A major privacy issue has come to light involving Grok, Elon Musk’s AI chatbot. Hundreds of thousands of user chats with the Grok chatbot have been found on Google search results. This happened without most users knowing their chats were being made public.

When users press a button to share their Grok chat, a unique link is created. The feature is meant to share the conversation with someone directly. But it also seems that these links were indexed by Google. That means anyone could find them through a simple search.

Grok Users Shocked as Hundreds of Thousands of Chats Appear in Google Search

Reports suggest that nearly 300,000 Grok chats were visible. Forbes, a tech publication, claimed the number could be even higher at around 370,000. This has led experts to raise an alarm about the risks of AI chatbots. One researcher even called them a “privacy disaster in progress.”

The BBC checked some of the exposed chats. They included people asking Grok to generate secure passwords, suggest diet plans, or even answer medical questions. In other cases, users tried pushing the chatbot’s limits with sensitive or dangerous prompts. Shockingly, one chat showed Grok giving instructions on how to make a Class A drug.

This is not the first time something like this has happened. OpenAI faced a similar situation when its ChatGPT “share” option allowed conversations to appear in search results. At the time, OpenAI said the chats were private by default. Users had to opt in to share them. The company later adjusted the feature to give people more control.

See Also: Apple App Store Bias? Elon Musk Claims X and Grok Face Unfair Rankings

Meta has also faced criticism. Earlier this year, shared conversations with its chatbot, Meta AI, appeared in a public feed inside the app. That meant users’ private queries could be seen by others.

Experts warn that even when names and accounts are hidden, chatbot conversations can still reveal a lot. Prompts might include personal details like health conditions, relationships, or even business information. Once these chats are out in the open, they cannot easily be removed. “Once leaked online, these conversations will stay there forever,” said Prof. Luc Rocher from Oxford Internet Institute.

Carissa Veliz, another professor at Oxford University, said the problem lies in users not being told what happens to their data. According to her, technology should be transparent, but in many cases it is not. She said, “Our technology doesn’t even tell us what it’s doing with our data, and that’s a problem.”

The incident raises bigger questions about AI safety. People trust chatbots with private information, often without thinking twice. But if these conversations are exposed, the consequences can be serious. Sensitive details could be misused by hackers, marketers, or even criminals.

At the moment, X, the parent company behind Grok, has not given a response. But the incident is a reminder for users to be careful. Before pressing “share,” it is important to understand where that data might end up. With AI becoming part of daily life, privacy safeguards must be stronger than ever.