An independent security researcher has uncovered a major data breach affecting Chat & Ask AI, one of the most popular AI chat applications on Google Play and Apple App Store with over 50 million users, according to a report from Malwarebytes. The researcher accessed 300 million messages from over 25 million users through an exposed Firebase database — a well-known and entirely preventable misconfiguration.
Scope of the Exposure
Chat & Ask AI functions as a "wrapper" application that connects users to multiple large language models including OpenAI's ChatGPT, Anthropic's Claude, and Google's Gemini. Users can select which model to interact with, making it a popular single-interface for AI conversations. The exposed database contained complete chat histories, the models used for each conversation, user settings, and critically — messages that reportedly included discussions of illegal activities and requests for suicide assistance.
The breach also revealed data belonging to users of other applications developed by Codeway, the parent company behind Chat & Ask AI, expanding the impact beyond a single app's user base.
A Systemic Problem: Firebase Misconfigurations
The vulnerability behind the breach is a well-documented Firebase misconfiguration where Security Rules are left set to public, allowing anyone with the project URL to read, modify, or delete data without authentication. The researcher, identified only as Harry, subsequently built an automated tool to scan apps for this vulnerability — and found that 103 out of 200 iOS apps scanned suffered from the same issue, collectively exposing tens of millions of stored files.
The incident underscores a troubling pattern: as AI wrapper applications proliferate across mobile platforms, many are built by developers who prioritize rapid market entry over security fundamentals. The combination of highly sensitive conversational data — users frequently share personal, medical, legal, and financial information with AI chatbots — and basic infrastructure misconfigurations creates catastrophic exposure scenarios.
Organizations building or evaluating AI applications should audit all backend-as-a-service configurations, implement authentication requirements on all database endpoints, conduct regular automated security scanning of mobile applications, and recognize that AI conversation data carries elevated sensitivity requiring enhanced data protection controls.