Artificial Intelligence and Consumer Apps

An independent security researcher says Chat & Ask AI, a mobile chatbot app with more than 50 million downloads on Apple’s App Store and Google Play, left roughly 300 million messages from over 25 million users exposed because of a misconfigured Google Firebase backend. By exploiting the error, the researcher was able to view complete chat histories, timestamps, the custom names users gave their bots, and which underlying AI models (ChatGPT, Claude, Gemini and others) were used—data that included highly sensitive content such as suicide plans, drug‑manufacturing questions and hacking guidance. Because Chat & Ask AI is just a 'wrapper' on top of large language models from OpenAI, Anthropic and Google, it—not the model providers—was storing this material, and experts note that the kind of Firebase misconfiguration involved is a well‑known and easily discoverable security weakness. The app’s publisher, Codeway, did not respond to Fox’s questions before publication, and there is no indication yet of how long the data was exposed or whether criminals accessed and scraped it. The incident underscores how many Americans treat AI chats like private journals or therapy sessions and how insecure third‑party apps can turn those conversations into a permanent trove for blackmail, doxxing or other abuse if developers fail basic cloud‑security hygiene.