Facts
Security researchers from Cybernews, a news portal and research company for cybersecurity, found an inadequately secured database at the AI app "Video AI Art Generator & Maker." The app is used for image and video editing. Due to a misconfigured Google Cloud Storage bucket, the media content with a total volume of twelve terabytes was publicly and accessible without authentication.
Affected were both more than 1.5 million images uploaded by users and more than 385,000 videos. In addition, around three million AI-generated videos and images and more than 386,000 AI-generated audio files were affected.
The affected app was listed in the Google Play Store with over 500,000 downloads and an average rating of 4.3 (out of 5) and was removed in the meantime. The developer, based in Turkey, had already previously stood out according to security researchers due to misconfigured backend services.
Assessment
AI-supported apps and tools for private use are gaining in popularity. At the same time, security incidents are also piling up in which image material, location data, or other sensitive information is inadequately protected or disclosed. In particular, there is a risk of AI-supported identity misuse, for example in the context of social engineering attacks or deepfake scenarios.
Reaction
Those affected are recommended to uninstall the app immediately and to remove any remaining app residues on the end devices. In addition, increased vigilance is warranted with regard to potential abusive use of the possibly exfiltrated image and video content. If corresponding cases of misuse become known, a report should be filed immediately with the competent law enforcement authorities.