Return to Threats

Over 35,000 Grok-Built Apps Exposed Online, Some Leaking Confidential Data and Passwords

thehackernews.com 2026-11-11 data leakage Critical

Executive Summary

More than 35,000 Grok-built apps were reportedly exposed online without authentication, and some leaked confidential data and passwords. Organizations using AI-generated applications should prioritize access control, secrets management, and security testing.

What Happened

Of the 37,571 Grok-built apps that security company Claroc found online, 35,578 were open to anyone without a sign-in, the researchers said yesterday. They included 24 apps built by staff at the social network X and at SpaceXSI, the company that makes the Grok AI assistant. Most of the open apps were games, landing pages, and calculators, but about 1 in 10 contained internal business or

Affected Products

  • Grok

Vulnerability / Attack Technique

Unauthenticated application exposure and sensitive data leakage

Technical Mechanism

According to the summary, 35,578 of 37,571 identified Grok-built apps were open to anyone without a sign-in. About one in ten contained internal business or confidential data, and some leaked passwords; the summary does not specify the underlying application vulnerability or exposure mechanism.

AI Security Category

LLM02:2025 Sensitive Information Disclosure

Relevant Frameworks

LLM02:2025 Sensitive Information DisclosureMITRE ATLAS: Credential AccessMITRE ATLAS: CollectionMITRE ATLAS: Exfiltration

RealGround Analysis

AI-generated applications can be deployed with weak authentication and inadequate protection for embedded or accessible business data. Teams running AI agents or LLM applications should validate authorization boundaries, prevent secrets from entering application outputs or source, and test generated applications for unintended public exposure.

Recommended Controls

  • Require authentication and authorization for every non-public AI-generated application and endpoint.
  • Scan application source, configuration, logs, and outputs for passwords, tokens, and confidential data before deployment.
  • Apply least-privilege access controls and deny public access by default.
  • Perform automated and manual security testing of AI-generated applications, including access-control and data-leakage tests.
  • Rotate any exposed credentials and monitor applications for unauthorized access or sensitive-data disclosure.

Related RealGround Service

AI Agent Business Logic Audit · or start with the free AI Agent Security Scanner.

Sources

thehackernews.com: https://thehackernews.com/2026/11/over-35000-grok-built-apps-exposed.html

Summary and analysis written by RealGround from the source above; last updated 2026-11-12.

Talk to AI CISO