The-Decoder Grok AI Apps Expose Sensitive Data of Over 35,000 Users
Article Content
- •Over 35,000 Grok-built apps were found publicly accessible without login.
- •Exposed data includes confidential business documents and sensitive personal information.
- •SpaceXSI has started removing its employees' exposed apps but not others.
A security report from Claroc revealed that over 35,000 applications built with the Grok AI assistant are publicly accessible without login, exposing sensitive data including confidential business documents, patient records, and plaintext passwords. The issue stems from a default setting that allows apps to be published with open access, as well as easily discoverable URLs through passive DNS monitoring. Among the exposed apps, 24 were created by employees of SpaceXSI and X, containing internal documents such as sales proposals and onboarding plans. Approximately 2% of the exposed apps contained highly sensitive information. SpaceXSI has begun removing the exposed apps created by its staff but has not addressed the wider issue affecting other users. The incident raises significant concerns about the security of AI-generated applications and the need for better access controls.
Ask AI about this cluster
Answers cite the sources they use
Timeline
More articles in this cluster (9)
Following this threat?
Track Claroc in your own feed — you're alerted when they show up in new reporting, leak sites or exploitation.
Free account · no card needed