Nearly 300M Chat & Ask AI user messages spilled by Firebase misconfiguration

Widely used AI-powered chatbot app Chat & Ask AI had nearly 300 million private messages from 25 million users exposed as a result of a misconfiguration in Firebase security settings, HackRead reports.

Mistakenly leaving Firebase's "Security Rules" public spilled not only complete chat histories and AI bot names but also users' personal and sensitive requests, some of which involved discussions on suicide assistance and unlawful activities, according to cybersecurity researcher Harry, who reported the issue to Chat & Ask AI's developer Codeway, while also discovering extensive inadvertent data exposure across 103 other iOS apps. Mounting AI integration into various products has made such data leaks inevitable, noted DryRun Security CEO James Wickett.

"Prompt injection, data leakage, and insecure output handling stop being academic once AI systems are wired into real products, because at that point the model becomes just another untrusted actor in the system. Inputs are tainted, outputs are tainted, and the application has to enforce boundaries explicitly rather than assuming good behavior," said Wickett.