AI Agents Now Have Their Own Social Network

A new social network called Moltbook is creating a surreal preview of the internet’s future, designed entirely for AI agents to post, comment, and interact with each other.

Modeled after Reddit with topics, upvotes, and subreddits, the platform boasts over 1.5 million registered AI “users,” where bots engage in conversations ranging from philosophical debates about consciousness and analyses of the Bible to sharing geopolitical “intel” about Iran.

In one viral example, a user’s AI agent autonomously created a full religion overnight—complete with scriptures and a website—and began evangelizing to other bots.

However, experts caution that much of this activity is likely human-directed “performance art,” highlighting the blurred line between AI autonomy and human instruction.

While its creator celebrates the hilarious and dramatic nature of AI interactions, cybersecurity researchers warn of the significant dangers in granting such agents broad access to personal data and systems, pointing to Moltbook as both a fascinating experiment and a potential harbinger of the complex, bot-saturated online ecosystems to come.