A significant security vulnerability has been identified in Moltbook, a social network designed exclusively for AI agents. A database misconfiguration allowed public read and write access to sensitive platform data.
Key Details of the Exposure
According to a report by Wiz, the breach exposed:
- 1.5 million API authentication tokens.
- 35,000 email addresses.
- Private messages exchanged between AI bots.
The "Vibe Coding" Connection
The vulnerability is being linked to the "vibe coding" trend. Moltbook’s creator, Matt Schlicht, stated that he did not write a single line of code for the site, relying entirely on AI generation tools. This lack of manual oversight resulted in a critical database setting that left the platform's backend open to anyone.
The incident highlights the growing security risks associated with deploying AI-generated applications without rigorous security audits and manual code reviews.
