Firehound analyzes app store apps with serious data leaks

Firehound has exposed a structural problem in the Apple App Store. Hundreds of apps are revealing sensitive user data, often without this being immediately apparent. Names, email addresses, and complete chat histories are freely accessible. Apps with AI features are particularly affected. Firehound collects, organizes, and documents these cases, revealing the true extent of the problem.

The project is led by the security research lab CovertLabs. Firehound became publicly known through the X user vxunderground, who drew attention to the repository and drastically summarized the situation. Firehound was developed by OSINT researcher Harrris0n, who took on the task of systematically collecting and analyzing problematic app store entries.

What exactly does Firehound do?

Firehound scans and indexes iOS apps that expose user data via insecurely configured databases or cloud storage. The project documents which data is affected, how many data records are accessible, and the underlying technical structures.

At the time of the analysis, 198 iOS apps were identified that disclose user information in some form. 196 of these apps specifically reveal user data. In many cases, the data is accessible without authentication.

Focus on AI apps

A clear focus is on AI-related applications. Many of these apps process text input, chats, or generated content and store it permanently. This is precisely where massive security vulnerabilities become apparent.

The app "Chat & Ask AI" leads the internal Firehound rankings for both the most exposed files and the most datasets. The numbers are clear:

more than 406 million data records
Data from over 18 million users

Among other things, chat histories, email addresses and other personal information will be disclosed.

How the data becomes public

Most of the affected apps use poorly secured cloud infrastructures. These often involve open databases or misconfigured storage solutions. Firehound not only shows that data is accessible, but often also:

the underlying data schemas
the exact number of data records
the type of information stored

This makes it clear how easily third parties can access this data.

Affected app categories

Although many entries have an AI connection, numerous other categories are affected:

Education
Entertainment
Graphic design
Health and Fitness
Lifestyle
Social networks
Other applications

This problem is therefore not a niche issue, but rather a problem that permeates the entire App Store.

Limited access to Firehound

Firehound does not make its results fully available to the public. Access to the public area is intentionally limited. Registration is required to access detailed scan results and restricted datasets. The process is clearly defined:

Access must be requested.
Each request is manually reviewed.
Priority will be given to journalists, law enforcement agencies and security experts.

Some of the data is so sensitive that it will only be released after responsible review and editing.

No confirmed evidence for „AI Slop“

Although it was initially claimed that Firehound specifically cataloged so-called "AI Slop," this term does not appear on either the Firehound website or Harrris0n's public profile. While many apps do use AI, it is currently impossible to definitively say whether they were created with Vibe Coding or other autonomous AI development tools.

Firehound makes the extent of data privacy problems visible

Firehound demonstrates the extent of the data privacy problem in the App Store. Millions of user data records are exposed, often due to simple configuration errors. AI apps, in particular, collect vast amounts of sensitive information and fail to adequately protect it. The project makes it clear that low barriers to entry in app development do not absolve developers of the responsibility to protect user data. Firehound is therefore less a scandal project than a reflection of a widespread problem. (Image: Shutterstock / Ar_TH)