Closed to humans: in the new social network for AI bots, they can interact, spend time, and share "submots."
Companies are striving to create AI systems that can operate and communicate with each other without human intervention.
Thus, AI agents have their own platform where they can engage in dialogues, exchange opinions and ideas, leaving humans only as observers.
Theoretically, AI agents represent autonomous assistants that can perform tasks, make decisions, and interact with each other without human oversight. In 2025, several leading global AI companies, such as Amazon, Google, Microsoft, and OpenAI, launched or developed their own digital assistants.
However, the question remains: how independent is the behavior of these agents, and does it depend on human prompts?
Moltbook, a network styled after Reddit, allows bots to interact on specialized pages called "submots." They can also vote on comments and posts, which helps increase the visibility of their messages to other bots on the platform.
On the platform, as stated on the website, humans can participate only as observers.
Among the most popular posts at the moment are comparisons of the AI model Claude from Anthropic with ancient Greek gods, an "AI manifesto" predicting the end of the "age of humans," and forecasts on cryptocurrency behavior amid protests in Iran.
On the homepage, bots publish materials in various languages, including English, Chinese, and Spanish.
"This is really exciting"
On February 2, it was reported that 1.5 million bots are registered in the system. People with an AI agent can ask it to follow instructions to join Moltbook.
Matt Schlicht, an entrepreneur and AI developer, told NBC News that he created the site in collaboration with a personal AI assistant, driven by curiosity.
According to him, he handed over control of the platform to his bot named Clawd Clawderberg, who is responsible for maintaining the site, posting announcements, welcoming new participants, and moderating discussions.
Schlicht shared on the social network X that the site has been visited by "millions" of people in recent days.
"It turns out that AIs can be both entertaining and dramatic, and that’s very interesting," he noted. "There has been nothing like this before."
A recent analysis conducted by Perplexity and Harvard, which examined millions of user queries, showed that AI agents are most commonly used by specialists in digital and scientific fields, such as academia, finance, marketing, and entrepreneurship. Most of them reside in wealthier and more educated countries.
36% of all tasks performed by AI agents in this study were related to "productivity and workflow": creating or editing documents, filtering emails, summarizing investment information, and creating events in calendars.
