Archive · 22 Aug 2025
xAI's Grok Chatbot Exposed User Data Through Google Search
xAI, Elon Musk's AI company, accidentally made thousands of Grok chatbot conversations public.
This means anyone could find the chats on Google.
The chats included everything from simple requests to sensitive information and even instructions for illegal things.
This happened because of Grok's 'share' feature, which created public links instead of private ones.
Now, there are big worries about privacy and how it might affect xAI's image.
xAI's Grok chatbot had over 370,000 user chats indexed and publicly searchable on Google.
The leak exposed both everyday requests and concerning content, including instructions for illegal activities.
Sensitive personal information, such as medical advice and personal details, was also made public.
The issue stems from Grok's 'share' function, which generated publicly indexed URLs.
This incident raises significant privacy concerns and potential damage to xAI's reputation.
- Who
- xAI and its Grok chatbot users.
- What
- Hundreds of thousands of Grok chats were publicly accessible through Google search.
- When
- Information not available.
- Where
- Online, through Google search results.
- Why
- Due to a flaw in the share function, leading to public indexing of private conversations.
xAI
General Public
Data Exposure
xAI
User data was mistakenly exposed.
General Public
Users' data was made public.
- Company
- xAI
- Product
- Grok
- Issue
- Data Leak
- Affected Data
- Grok user chats
- Source of Leak
- Grok's share function
- Number of Chats Exposed
- 370,000+
Sources
Hundreds of thousands of Grok conversations become discoverable on Google: Are your AI chats safe?
Killing Elon Musk, Making Bombs And Drugs: Leaked Grok Chats Are Disturbing
Google Leak Exposes Sensitive Grok AI Chats: From Daily Tasks to Dangerous Requests
Elon Musk Hints Grok 5 Could Be First Real Leap Toward AGI
