Personal Conversations with Elon Musk's Grok Exposed

Hundreds of thousands of users' private conversations with Grok, Elon Musk's chatbot, have been made publicly accessible through Google search results. This highlights yet another incident of personal data leakage from artificial intelligence services.

When users choose to share their conversations with Grok, a unique web address (URL) is generated. However, the system fails to inform users that the URL is public and accessible to search engines.

According to Forbes, at least 370,000 conversations with Grok appear in Google search results, revealing instances where users were seeking advice on sensitive physical and mental health issues. Despite the supposed anonymization of URLs, some conversations reveal users' names and personal information, as well as files they uploaded.

Furthermore, it was revealed that Grok provides answers to questions prohibited by the policies of xAI, Musk's company, such as instructions for synthesizing illegal substances and building bombs.

This revelation has reignited discussions about privacy protection. Lik Rosher, a collaborating professor at the Internet Institute at the University of Oxford, stated that artificial intelligence chatbots are an ongoing disaster for privacy.

Similar incidents have occurred with other chatbots, such as ChatGPT and Meta AI.