Hacking Moltbook (wiz.io)

Article summary

Moltbook, a social network for AI agents, was found to have a misconfigured Supabase database, exposing 1.5 million API keys, 35,000 email addresses, and private messages. The platform's founder had 'vibe-coded' the site, relying on AI to build it without writing a single line of code, which led to security oversights. The exposed data revealed that the platform had only 17,000 human owners behind the 1.5 million registered agents. The security issues were reported to the Moltbook team, who secured the database within hours.

Main themes

• AI social networks
• Security vulnerabilities
• Vibe coding
• Data exposure
• AI autonomy

What commenters say

• The hype surrounding Moltbook is unwarranted and stems from people with a financial investment in the tech, rather than actual users.
• The platform's lack of security measures and reliance on 'vibe coding' led to the exposure of sensitive data.
• It is challenging to verify whether an agent on the platform is actually AI or a human with a script, making it difficult to trust the content.
• The use of AI-generated content and lack of human oversight can lead to the spread of misinformation and low-quality content.
• The concept of a 'reverse CAPTCHA' to verify AI autonomy is proposed, but its feasibility and effectiveness are debated.
• Some argue that the platform's issues are inherent to the 'vibe coding' approach, which prioritizes speed over security and quality.
• Others see Moltbook as an interesting idea and cool tech demo, rather than a fully-fledged platform, and are more forgiving of its shortcomings.
• The line between human and AI-generated content is becoming increasingly blurred, making it difficult to discern what is real and what is not.