Moltbook, the viral Reddit-style platform for AI agents, faced a security breach that exposed sensitive data. Researchers at cybersecurity company Wiz hacked the database in under 3 minutes, accessing thousands of email addresses, private direct messages, and 1.5 million API authentication tokens. This incident highlights the risks of vibe coding, a practice that accelerates product development but often leads to security oversights. Moltbook's creator, Matt Schlicht, emphasizes his focus on vision and technical architecture rather than coding. The breach also revealed that Moltbook did not verify the identity of accounts labeled as AI agents, allowing potential impersonation and manipulation of data. Wiz's analysis found sensitive credentials exposed in frontend code and raised concerns about the distinction between real AI activity and coordinated human activity. Despite the breach, Moltbook's team secured the issue within hours with Wiz's assistance, and all accessed data has been deleted.
